
import os
import sys
import math
import scipy as scipy
import numpy as np
from numpy import arange
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import cv2
import ast
import pathlib

import plotly.express as px
import plotly.graph_objects as go
import hashlib
from IPython.display import display
from PIL import Image
from tqdm import tqdm
tqdm.pandas()

%matplotlib inline


# Setting the display option
pd.set_option('display.max_columns', 50)

# Defining the variables for better plotting 
label_font_dict = {'family':'sans-serif','size':13.5,'color':'brown','style':'italic'}
title_font_dict = {'family':'sans-serif','size':17.5,'color':'Blue','style':'italic'}


!pip install tensorflow-addons

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting tensorflow-addons
  Downloading tensorflow_addons-0.18.0-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.1 MB)
     |████████████████████████████████| 1.1 MB 30.7 MB/s 
Requirement already satisfied: typeguard>=2.7 in /usr/local/lib/python3.8/dist-packages (from tensorflow-addons) (2.7.1)
Requirement already satisfied: packaging in /usr/local/lib/python3.8/dist-packages (from tensorflow-addons) (21.3)
Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.8/dist-packages (from packaging->tensorflow-addons) (3.0.9)
Installing collected packages: tensorflow-addons
Successfully installed tensorflow-addons-0.18.0


## TRAIN set files
!gdown 1kuGJodL48MaAb57j2VSbSX_qXellGrBC
!gdown 1NAwb73kvK_uut71vNIte34tiaDKfSmTZ

Downloading...
From: https://drive.google.com/uc?id=1kuGJodL48MaAb57j2VSbSX_qXellGrBC
To: /content/Dataset.zip
100% 203k/203k [00:00<00:00, 110MB/s]
Downloading...
From: https://drive.google.com/uc?id=1NAwb73kvK_uut71vNIte34tiaDKfSmTZ
To: /content/HP_data_engg.csv
100% 3.55k/3.55k [00:00<00:00, 6.40MB/s]


!unzip Dataset.zip

Archive:  Dataset.zip
  inflating: Dataset/1d5fee00-ada4-11eb-a80b-3f50af21830f.jpg  
  inflating: Dataset/1eddf4c0-ad9f-11eb-a80b-ef4adba049aa.jpg  
  inflating: Dataset/38e29bc0-8fc0-11eb-b5f2-0742da91b282.jpg  
  inflating: Dataset/6cd5d480-8b8b-11eb-b152-f7b115384fa9.jpg  
  inflating: Dataset/76ed6fc0-8e1d-11eb-ad70-7be1a1219834.jpg  
  inflating: Dataset/8d23fa60-9201-11eb-b3e6-cbfb48c6a300.jpg  
  inflating: Dataset/9dc3eba0-8d8d-11eb-8d3f-4776be867c61.jpg  
  inflating: Dataset/9dd020a0-8d8d-11eb-8d3f-bfd9f6bd04d4.jpg  
  inflating: Dataset/d3757800-8dff-11eb-a328-472d4f8db277.jpg


# Storing the location of the 'height_and_pose' file
hp_df = pd.read_csv(os.path.join(os.getcwd(), "HP_data_engg.csv"), index_col="Unnamed: 0")
hp_df['Img_name'] = hp_df['Img_name'].apply(lambda name: os.path.join(os.path.join(os.getcwd(), "Dataset"), name+".jpg"))
hp_df


hp_df.shape

(9, 36)


import random as rn
import datetime
import time
import tensorflow as tf
import tensorflow_addons as tfa
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense, Input, Dropout, BatchNormalization, Activation, Flatten, Conv1D, Conv2D, Concatenate, Lambda
from tensorflow.keras.optimizers import Adam, RMSprop
from tensorflow.keras.utils import plot_model
from tensorflow.keras.layers import MaxPool1D, MaxPool2D, GlobalAveragePooling1D, GlobalAveragePooling2D, AveragePooling1D, AveragePooling2D, SpatialDropout2D
from keras.regularizers import l1, l2
from keras.preprocessing import image
from keras.callbacks import Callback
from keras.callbacks import TensorBoard
from sklearn.metrics import accuracy_score, recall_score, f1_score, precision_score, roc_auc_score, confusion_matrix
from tensorboard import notebook
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score


# All train images paths
all_train_paths = hp_df['Img_name'].values

# All train images tgt values
all_train_labels = np.float32(hp_df['Height'].values)

# Train and Validation Split
train_paths, valid_paths, train_labels, valid_labels = train_test_split(all_train_paths, 
                                                                        all_train_labels, 
                                                                        test_size=0.20, 
                                                                        random_state=2022)


# Checking shapes of all images paths and tgt values
all_train_paths.shape, all_train_labels.shape

((9,), (9,))


# Checking shapes of all TRAIN & VALIDATON images paths and tgt values
train_paths.shape, train_labels.shape, valid_paths.shape, valid_labels.shape

((7,), (7,), (2,), (2,))


# Function to read all the images
def reading_images(imgs_paths, size=(448,448)):
    """
    Description : This function will read all the given images and store them in a list.
    """
    # List container for storing all the images after reading, resizing & normalization
    images = []
    for index, img_path in enumerate(imgs_paths):
        # Reading the image
        img = cv2.imread(img_path)
        
        # Changing the color to RGB model
        img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
        
        # Resizing the image
        img = cv2.resize(img, size)
        
        # Normalizing the pixel values
        img = img / 255.0
        
        # Storing every image
        images.append(img)
    return images


# Checking the train images paths
train_paths

array(['/content/Dataset/76ed6fc0-8e1d-11eb-ad70-7be1a1219834.jpg',
       '/content/Dataset/38e29bc0-8fc0-11eb-b5f2-0742da91b282.jpg',
       '/content/Dataset/8d23fa60-9201-11eb-b3e6-cbfb48c6a300.jpg',
       '/content/Dataset/d3757800-8dff-11eb-a328-472d4f8db277.jpg',
       '/content/Dataset/9dd020a0-8d8d-11eb-8d3f-bfd9f6bd04d4.jpg',
       '/content/Dataset/1eddf4c0-ad9f-11eb-a80b-ef4adba049aa.jpg',
       '/content/Dataset/1d5fee00-ada4-11eb-a80b-3f50af21830f.jpg'],
      dtype=object)


# Reading all the TRAIN images in a single list
all_train_images = reading_images(imgs_paths=train_paths)

# Matching the count of existing TRAIN images before-after reading them
print("### Total TRAIN Images & Labels available after splitting are  ===>  {} and {} ###".format(len(all_train_images),
                                                                                                   train_labels.shape[0]))

### Total TRAIN Images & Labels available after splitting are  ===>  7 and 7 ###


# Checking the shape of train image after reading it
print("### The shape of train image after reading it is  ===>   {}   ###".format(all_train_images[0].shape))

### The shape of train image after reading it is  ===>   (448, 448, 3)   ###


# Reading all the TEST images in a single list
all_test_images = reading_images(imgs_paths=valid_paths)

# Matching the count of existing TEST images before-after reading them
print("### Total TEST Images & Labels available after splitting are  ===>  {} and {} ###".format(len(all_test_images),
                                                                                                   valid_labels.shape[0]))

### Total TEST Images & Labels available after splitting are  ===>  2 and 2 ###


# Checking the shape of test image after reading it
print("### The shape of test image after reading it is  ===>   {}   ###".format(all_test_images[0].shape))

### The shape of test image after reading it is  ===>   (448, 448, 3)   ###


# Storing X_test and y_test for evaluation of models
X_test = np.asarray(all_test_images)
y_test = valid_labels

# Checking the dtype & shape of test set after reading it
print("### The dtype & shape of test set after reading it is  ===>   {}, {}, {}, {}  ###".format(type(X_test), type(y_test), X_test.shape, y_test.shape))

### The dtype & shape of test set after reading it is  ===>   <class 'numpy.ndarray'>, <class 'numpy.ndarray'>, (2, 448, 448, 3), (2,)  ###


# Checking the shape of the tgt values
valid_labels.shape

(2,)


# Defining the directory for storing the augmented images
aug_imgs_path = pathlib.Path(os.path.join(os.path.join(os.getcwd(),"Dataset"), "Augmented_Images"))
aug_imgs_path

PosixPath('/content/Dataset/Augmented_Images')


# Creating the folder in the drive for storing the Augmented Images
if "Augmented_Images" not in os.listdir(os.path.join(os.getcwd(),"Dataset")):
    os.mkdir(aug_imgs_path)
else:
    None


def data_augmentation(image):
    """
    Description : This function is created for performing the augmentations on an image to yield a different image.
    
    Inputs : It accepts below parameter:
        - img_name : Image name
    
    Returns : It returns the tensor object of an augmented image.
    """     
    # Adjusting the hue of the RGB image
    image = tf.image.adjust_hue(image, delta=0.02)
    
    # Adjusting the Brightness of the RGB image
    image = tf.image.adjust_brightness(image, delta=0.05)
   
    # Reducing the noise using Gaussian filter
    image = tfa.image.gaussian_filter2d(image, filter_shape=(12,12))
       
    return image


# Defining the IMAGE DATA GENERATOR for only the TRAIN SET
train_img_gen = tf.keras.preprocessing.image.ImageDataGenerator(rotation_range=0.3, 
                                                                zoom_range=0.08,
                                                                horizontal_flip=True, 
                                                                vertical_flip=False, 
                                                                shear_range=0.25, 
                                                                width_shift_range=0.15,
                                                                height_shift_range=0.15,
                                                                preprocessing_function=data_augmentation,
                                                                rescale=None)


def generate_aug_images(readed_images, labels, data_gen, aug_img_dir, size=(448,448), no_of_aug_images=9):
    """
    Description : This function is created for generating & storing(both in-drive & memory) the augmented images.
    """
    images_after_aug = []
    labels_after_aug = []
    for index, image in enumerate(readed_images):
        for i in range(no_of_aug_images):
            img = data_gen.flow(x=np.reshape(image, (1, size[0], size[1], 3)), 
                                  batch_size=32,
                                  shuffle=True,
                                  sample_weight=None,
                                  save_to_dir=aug_imgs_path,
                                  save_prefix="Aug_Train",
                                  save_format='jpg').next()
            images_after_aug.append(np.reshape(img, (size[0], size[1], 3)))
            labels_after_aug.append(train_labels[index])

    aug_labels = np.asarray(labels_after_aug)
    aug_images = np.asarray(images_after_aug)
    return aug_images, aug_labels


# Generating the augmented images of only 1 image
augmented_images1, augmented_labels1 = generate_aug_images(all_train_images[0:1], 
                                                         labels=train_labels, 
                                                         data_gen=train_img_gen, 
                                                         aug_img_dir=aug_imgs_path)


# Checking the shapes of augmented images & labels array
augmented_images1.shape, augmented_labels1.shape

((9, 448, 448, 3), (9,))


# Checking the image of 0th index augmented image
augmented_images1[0][0].shape

(448, 3)


def plot_aug_images(org_images, aug_images):
    """
    Description : This function is created for visualizing the original and its augmented images.
    """
    with plt.style.context('seaborn'):
        fig, ax = plt.subplots(nrows=2, ncols=5, figsize=(16,8), sharex=True, sharey=True)
        ax[0,0].imshow(org_images[0], aspect='auto')
        ax[0,0].set_title("Original", fontdict= label_font_dict)
        ax[0,0].axis("off")
        r = 0;
        c = 1;
        for i in range(9):
            ax[r,c].imshow(aug_images[i], aspect='auto')
            ax[r,c].set_title("Aug Image - {}".format(i+1), fontdict= label_font_dict)
            ax[r,c].axis("off")
            c +=1;
            if c >= 5:
                r = 1;
                c = 0;
        plt.show();


# Lets see the augmented images
plot_aug_images(org_images=all_train_images, aug_images=augmented_images1)


# Generating the augmented images of only 1 image
augmented_images2, augmented_labels2 = generate_aug_images(all_train_images[2:3], 
                                                         labels=train_labels, 
                                                         data_gen=train_img_gen, 
                                                         aug_img_dir=aug_imgs_path)


# Checking the shapes of augmented images & labels array
augmented_images2.shape, augmented_labels2.shape

((9, 448, 448, 3), (9,))


# Checking the image of 0th index augmented image
augmented_images2[0][0].shape

(448, 3)


# Lets see the augmented images
plot_aug_images(org_images=all_train_images[2:3], aug_images=augmented_images2)


# Generating the augmented images of only 1 image
augmented_images3, augmented_labels3 = generate_aug_images(all_train_images[6:7], 
                                                         labels=train_labels, 
                                                         data_gen=train_img_gen, 
                                                         aug_img_dir=aug_imgs_path)


# Checking the shapes of augmented images & labels array
augmented_images3.shape, augmented_labels3.shape

((9, 448, 448, 3), (9,))


# Checking the image of 0th index augmented image
augmented_images3[0][0].shape

(448, 3)


# Lets see the augmented images
plot_aug_images(org_images=all_train_images[6:7], aug_images=augmented_images3)


# Generating the augmented images of all train images
augmented_images, augmented_labels = generate_aug_images(all_train_images, 
                                                         labels=train_labels, 
                                                         data_gen=train_img_gen, 
                                                         aug_img_dir=aug_imgs_path)


# Checking the shapes of augmented images & labels array
augmented_images.shape, augmented_labels.shape

((63, 448, 448, 3), (63,))


# Checking the image of 0th index augmented image
augmented_images[0][0].shape

(448, 3)


# Train and Validation Split
X_train, X_val, y_train, y_val = train_test_split(augmented_images, augmented_labels, test_size = 0.15, random_state = 44)


# Shape of TRAIN & VALIDATION sets after augmentation
X_train.shape, X_val.shape, y_train.shape, y_val.shape

((53, 448, 448, 3), (10, 448, 448, 3), (53,), (10,))


# Checking the shape of the train tgt values object
y_train.shape

(53,)


# Checking the train tgt values object
y_train

array([ 76.7,  63.2,  63.2,  63.2,  63.2,  59.1,  63.2, 106.1,  99.9,
        66.4,  66.4,  99.9,  76.7, 106.1,  63.2,  99.9,  86.4,  66.4,
        66.4, 106.1,  99.9, 106.1,  59.1,  86.4,  99.9,  63.2,  76.7,
        86.4,  86.4,  99.9, 106.1, 106.1,  63.2,  59.1, 106.1, 106.1,
        76.7,  76.7,  59.1,  76.7,  86.4,  86.4,  86.4, 106.1,  66.4,
        76.7,  76.7,  99.9,  59.1,  66.4,  66.4,  99.9,  86.4],
      dtype=float32)


# Checking the shape of the validation tgt values object
y_val.shape

(10,)


# Checking the validation values object
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# Checking type of all the above objects
type(X_train), type(y_train), type(X_val), type(y_val)

(numpy.ndarray, numpy.ndarray, numpy.ndarray, numpy.ndarray)


# Saving the pre-processed TRAIN, VALIDATION & TEST sets
np.savez("X_train.npz",X_train)
np.savez("y_train.npz",y_train)
np.savez("X_val.npz",X_val)
np.savez("y_val.npz",y_val)
np.savez("X_test.npz",X_test)
np.savez("y_test.npz",y_test)


# mount the drive
from google.colab import drive
drive.mount('/content/drive')

Mounted at /content/drive


# copy it there
!cp X_train.npz /content/drive/MyDrive/CGM_Assignment
!cp y_train.npz /content/drive/MyDrive/CGM_Assignment
!cp X_val.npz /content/drive/MyDrive/CGM_Assignment
!cp y_val.npz /content/drive/MyDrive/CGM_Assignment
!cp X_test.npz /content/drive/MyDrive/CGM_Assignment
!cp y_test.npz /content/drive/MyDrive/CGM_Assignment


# Downloading the X_train, y_train, X_val, y_val, X_test, y_test files
!gdown 1524vFizdo4a0SuXIZJQITtxbQbW9Brgi
!gdown 1DzBoXkHsPTmQujHpZ4wTHY9BSb6jDsyv
!gdown 1NgXIJ_7yAsVMv2tqrpTKX3y6CinoarvY
!gdown 1O5le8A_ZQ1orlNqP2uE5dLfsXR8Wjkbq
!gdown 1Uu2EELSljqobSS1tH1xDNiJg22MRuiTq
!gdown 1woSkEH-jlC7Vbl4ZTHY_6F14PdCiJMUC


# Loading the X_train, y_train, X_val, y_val, X_test & y_test files in memory
x_train_file = np.load("X_train.npz")
X_train = x_train_file.f.arr_0

y_train_file = np.load("y_train.npz")
y_train = y_train_file.f.arr_0

x_val_file = np.load("X_val.npz")
X_val = x_val_file.f.arr_0

y_val_file = np.load("y_val.npz")
y_val = y_val_file.f.arr_0

x_test_file = np.load("X_test.npz")
X_test = x_test_file.f.arr_0

y_test_file = np.load("y_test.npz")
y_test = y_test_file.f.arr_0


# Checking types of TRAIN, VAL & TEST sets
type(X_train), type(y_train), type(X_val), type(y_val), type(X_test), type(y_test)

(numpy.ndarray,
 numpy.ndarray,
 numpy.ndarray,
 numpy.ndarray,
 numpy.ndarray,
 numpy.ndarray)


# Checking shapes of TRAIN, VAL & TEST sets
X_train.shape, y_train.shape, X_val.shape, y_val.shape, X_test.shape, y_test.shape

((53, 448, 448, 3), (53,), (10, 448, 448, 3), (10,), (2, 448, 448, 3), (2,))


from keras import backend as K

def r_square(y_true, y_pred):
    """
    Description : This function is created for calculating the value of the r-square.
    """
    # Calculating the residual
    diff = (y_true - y_pred)

    # Calculating the sum of squares of the residuals
    sum_of_squares_res = K.sum(K.square(diff)) 

    # Calculating the sum of squares of the total
    mean_y_true = K.mean(y_true)
    sum_of_squares_tot = K.sum(K.square(y_true - mean_y_true))

    # Calculating the R-Square
    r2_value = 1 - (sum_of_squares_res / (sum_of_squares_tot + K.epsilon()))
    return r2_value


# Loading the tensorboard
%load_ext tensorboard

# Folder storing all the runs logs
root_logdir = os.path.join(os.path.join(os.curdir, "logs"), "fit")

# Get current run logs directory
def get_run_logdir():
    run_id = time.strftime("run_%Y_%m_%d-%H_%M_%S")
    return os.path.join(root_logdir, run_id)


# Defining the batch-size
BATCH_SIZE = 9

# Using the AUTOTUNE algo for better allocation of computing resources for Dataset pipeline
AUTO = tf.data.experimental.AUTOTUNE


# build the ResNet-50 network
resnet50_with_no_top_model = tf.keras.applications.ResNet50(include_top=False, weights='imagenet', input_shape=(448,448,3))

Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5
94765736/94765736 [==============================] - 0s 0us/step


# Model summary
resnet50_with_no_top_model.summary()

Model: "resnet50"
__________________________________________________________________________________________________
 Layer (type)                   Output Shape         Param #     Connected to                     
==================================================================================================
 input_1 (InputLayer)           [(None, 448, 448, 3  0           []                               
                                )]                                                                
                                                                                                  
 conv1_pad (ZeroPadding2D)      (None, 454, 454, 3)  0           ['input_1[0][0]']                
                                                                                                  
 conv1_conv (Conv2D)            (None, 224, 224, 64  9472        ['conv1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv1_bn (BatchNormalization)  (None, 224, 224, 64  256         ['conv1_conv[0][0]']             
                                )                                                                 
                                                                                                  
 conv1_relu (Activation)        (None, 224, 224, 64  0           ['conv1_bn[0][0]']               
                                )                                                                 
                                                                                                  
 pool1_pad (ZeroPadding2D)      (None, 226, 226, 64  0           ['conv1_relu[0][0]']             
                                )                                                                 
                                                                                                  
 pool1_pool (MaxPooling2D)      (None, 112, 112, 64  0           ['pool1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv2_block1_1_conv (Conv2D)   (None, 112, 112, 64  4160        ['pool1_pool[0][0]']             
                                )                                                                 
                                                                                                  
 conv2_block1_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block1_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block1_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_0_conv (Conv2D)   (None, 112, 112, 25  16640       ['pool1_pool[0][0]']             
                                6)                                                                
                                                                                                  
 conv2_block1_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block1_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block1_0_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_0_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_0_bn[0][0]',      
                                6)                                'conv2_block1_3_bn[0][0]']      
                                                                                                  
 conv2_block1_out (Activation)  (None, 112, 112, 25  0           ['conv2_block1_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block2_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block1_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block2_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block2_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block2_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block2_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block2_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block2_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block2_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_out[0][0]',       
                                6)                                'conv2_block2_3_bn[0][0]']      
                                                                                                  
 conv2_block2_out (Activation)  (None, 112, 112, 25  0           ['conv2_block2_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block3_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block2_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block3_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block3_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block3_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block3_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block3_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block3_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block3_add (Add)         (None, 112, 112, 25  0           ['conv2_block2_out[0][0]',       
                                6)                                'conv2_block3_3_bn[0][0]']      
                                                                                                  
 conv2_block3_out (Activation)  (None, 112, 112, 25  0           ['conv2_block3_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv3_block1_1_conv (Conv2D)   (None, 56, 56, 128)  32896       ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block1_1_relu[0][0]']    
                                                                                                  
 conv3_block1_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_0_conv (Conv2D)   (None, 56, 56, 512)  131584      ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block1_2_relu[0][0]']    
                                                                                                  
 conv3_block1_0_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_0_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_0_bn[0][0]',      
                                                                  'conv3_block1_3_bn[0][0]']      
                                                                                                  
 conv3_block1_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block1_add[0][0]']       
                                                                                                  
 conv3_block2_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block1_out[0][0]']       
                                                                                                  
 conv3_block2_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block2_1_relu[0][0]']    
                                                                                                  
 conv3_block2_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block2_2_relu[0][0]']    
                                                                                                  
 conv3_block2_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block2_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_out[0][0]',       
                                                                  'conv3_block2_3_bn[0][0]']      
                                                                                                  
 conv3_block2_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block2_add[0][0]']       
                                                                                                  
 conv3_block3_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block2_out[0][0]']       
                                                                                                  
 conv3_block3_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block3_1_relu[0][0]']    
                                                                                                  
 conv3_block3_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block3_2_relu[0][0]']    
                                                                                                  
 conv3_block3_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block3_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_add (Add)         (None, 56, 56, 512)  0           ['conv3_block2_out[0][0]',       
                                                                  'conv3_block3_3_bn[0][0]']      
                                                                                                  
 conv3_block3_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block3_add[0][0]']       
                                                                                                  
 conv3_block4_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block3_out[0][0]']       
                                                                                                  
 conv3_block4_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block4_1_relu[0][0]']    
                                                                                                  
 conv3_block4_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block4_2_relu[0][0]']    
                                                                                                  
 conv3_block4_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block4_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_add (Add)         (None, 56, 56, 512)  0           ['conv3_block3_out[0][0]',       
                                                                  'conv3_block4_3_bn[0][0]']      
                                                                                                  
 conv3_block4_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block4_add[0][0]']       
                                                                                                  
 conv4_block1_1_conv (Conv2D)   (None, 28, 28, 256)  131328      ['conv3_block4_out[0][0]']       
                                                                                                  
 conv4_block1_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block1_1_relu[0][0]']    
                                                                                                  
 conv4_block1_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_0_conv (Conv2D)   (None, 28, 28, 1024  525312      ['conv3_block4_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block1_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block1_0_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_0_bn[0][0]',      
                                )                                 'conv4_block1_3_bn[0][0]']      
                                                                                                  
 conv4_block1_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block2_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block1_out[0][0]']       
                                                                                                  
 conv4_block2_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block2_1_relu[0][0]']    
                                                                                                  
 conv4_block2_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block2_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block2_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_out[0][0]',       
                                )                                 'conv4_block2_3_bn[0][0]']      
                                                                                                  
 conv4_block2_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block3_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block2_out[0][0]']       
                                                                                                  
 conv4_block3_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block3_1_relu[0][0]']    
                                                                                                  
 conv4_block3_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block3_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block3_add (Add)         (None, 28, 28, 1024  0           ['conv4_block2_out[0][0]',       
                                )                                 'conv4_block3_3_bn[0][0]']      
                                                                                                  
 conv4_block3_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block4_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block3_out[0][0]']       
                                                                                                  
 conv4_block4_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block4_1_relu[0][0]']    
                                                                                                  
 conv4_block4_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block4_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block4_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block4_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block4_add (Add)         (None, 28, 28, 1024  0           ['conv4_block3_out[0][0]',       
                                )                                 'conv4_block4_3_bn[0][0]']      
                                                                                                  
 conv4_block4_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block4_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block5_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block4_out[0][0]']       
                                                                                                  
 conv4_block5_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block5_1_relu[0][0]']    
                                                                                                  
 conv4_block5_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block5_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block5_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block5_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block5_add (Add)         (None, 28, 28, 1024  0           ['conv4_block4_out[0][0]',       
                                )                                 'conv4_block5_3_bn[0][0]']      
                                                                                                  
 conv4_block5_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block5_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block6_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block5_out[0][0]']       
                                                                                                  
 conv4_block6_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block6_1_relu[0][0]']    
                                                                                                  
 conv4_block6_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block6_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block6_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block6_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block6_add (Add)         (None, 28, 28, 1024  0           ['conv4_block5_out[0][0]',       
                                )                                 'conv4_block6_3_bn[0][0]']      
                                                                                                  
 conv4_block6_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block6_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_1_conv (Conv2D)   (None, 14, 14, 512)  524800      ['conv4_block6_out[0][0]']       
                                                                                                  
 conv5_block1_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block1_1_relu[0][0]']    
                                                                                                  
 conv5_block1_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_0_conv (Conv2D)   (None, 14, 14, 2048  2099200     ['conv4_block6_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block1_0_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_0_bn[0][0]',      
                                )                                 'conv5_block1_3_bn[0][0]']      
                                                                                                  
 conv5_block1_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block2_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block1_out[0][0]']       
                                                                                                  
 conv5_block2_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block2_1_relu[0][0]']    
                                                                                                  
 conv5_block2_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block2_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block2_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_out[0][0]',       
                                )                                 'conv5_block2_3_bn[0][0]']      
                                                                                                  
 conv5_block2_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block3_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block2_out[0][0]']       
                                                                                                  
 conv5_block3_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block3_1_relu[0][0]']    
                                                                                                  
 conv5_block3_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block3_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block3_add (Add)         (None, 14, 14, 2048  0           ['conv5_block2_out[0][0]',       
                                )                                 'conv5_block3_3_bn[0][0]']      
                                                                                                  
 conv5_block3_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
==================================================================================================
Total params: 23,587,712
Trainable params: 23,534,592
Non-trainable params: 53,120
__________________________________________________________________________________________________


# Instantiating Optimizer
learning_rate= 0.0001
opt1 = tf.keras.optimizers.Adam(learning_rate=learning_rate)

# Reduce Learning Rate on Plateau
reduce_lr1 = tf.keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.1, patience=2, verbose=1, mode='auto', min_delta=0.0001)


# Logs directory
curr_run_logdir1 = get_run_logdir()

# Instantiating Tensorboard callback
tensorboard_callback1 = TensorBoard(log_dir=curr_run_logdir1, histogram_freq=1)


# Setting the seed
os.environ['PYTHONHASHSEED'] = '0'

# Clearing the TF session
tf.keras.backend.clear_session()

# defining the custom top of the ResNet-50 model
for layer in resnet50_with_no_top_model.layers[0:-3]:
    layer.trainable = False

# Adding additional layers
input_layer = resnet50_with_no_top_model.output
    
# Defining the top layers structure of the model
flatten = tf.keras.layers.GlobalAveragePooling2D(name='Flatten_for_hidden_layers')(input_layer)

dropout_1 = Dropout(rate=0.5, name='Dropout1')(flatten)

dense_layer1 = tf.keras.layers.Dense(units=128, 
                                     activation='relu', 
                                     use_bias=True, 
                                     kernel_initializer=tf.keras.initializers.he_normal(seed=80),
                                     bias_initializer=tf.keras.initializers.he_normal(seed=110),
                                     name='Hidden_Layer1')(dropout_1)

dropout_2 = Dropout(rate=0.5, name='Dropout2')(dense_layer1)

dense_layer2 = tf.keras.layers.Dense(units=128, 
                                     activation='relu', 
                                     use_bias=True, 
                                     kernel_initializer=tf.keras.initializers.he_normal(seed=80),
                                     bias_initializer=tf.keras.initializers.he_normal(seed=110),
                                     name='Hidden_Layer2')(dropout_2)

output_layer = tf.keras.layers.Dense(1, activation='linear', name="output")(dense_layer2)

# Instantiating the complete model
resnet_50 = Model(inputs=resnet50_with_no_top_model.input, outputs=output_layer)

# Compiling the model
resnet_50.compile(optimizer=opt1,
                  loss = 'mean_squared_error',
                  metrics=[r_square])

# Summary of the ResNet-50 model with custom top
resnet_50.summary()

Model: "model"
__________________________________________________________________________________________________
 Layer (type)                   Output Shape         Param #     Connected to                     
==================================================================================================
 input_1 (InputLayer)           [(None, 448, 448, 3  0           []                               
                                )]                                                                
                                                                                                  
 conv1_pad (ZeroPadding2D)      (None, 454, 454, 3)  0           ['input_1[0][0]']                
                                                                                                  
 conv1_conv (Conv2D)            (None, 224, 224, 64  9472        ['conv1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv1_bn (BatchNormalization)  (None, 224, 224, 64  256         ['conv1_conv[0][0]']             
                                )                                                                 
                                                                                                  
 conv1_relu (Activation)        (None, 224, 224, 64  0           ['conv1_bn[0][0]']               
                                )                                                                 
                                                                                                  
 pool1_pad (ZeroPadding2D)      (None, 226, 226, 64  0           ['conv1_relu[0][0]']             
                                )                                                                 
                                                                                                  
 pool1_pool (MaxPooling2D)      (None, 112, 112, 64  0           ['pool1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv2_block1_1_conv (Conv2D)   (None, 112, 112, 64  4160        ['pool1_pool[0][0]']             
                                )                                                                 
                                                                                                  
 conv2_block1_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block1_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block1_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_0_conv (Conv2D)   (None, 112, 112, 25  16640       ['pool1_pool[0][0]']             
                                6)                                                                
                                                                                                  
 conv2_block1_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block1_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block1_0_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_0_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_0_bn[0][0]',      
                                6)                                'conv2_block1_3_bn[0][0]']      
                                                                                                  
 conv2_block1_out (Activation)  (None, 112, 112, 25  0           ['conv2_block1_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block2_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block1_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block2_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block2_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block2_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block2_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block2_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block2_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block2_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_out[0][0]',       
                                6)                                'conv2_block2_3_bn[0][0]']      
                                                                                                  
 conv2_block2_out (Activation)  (None, 112, 112, 25  0           ['conv2_block2_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block3_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block2_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block3_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block3_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block3_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block3_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block3_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block3_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block3_add (Add)         (None, 112, 112, 25  0           ['conv2_block2_out[0][0]',       
                                6)                                'conv2_block3_3_bn[0][0]']      
                                                                                                  
 conv2_block3_out (Activation)  (None, 112, 112, 25  0           ['conv2_block3_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv3_block1_1_conv (Conv2D)   (None, 56, 56, 128)  32896       ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block1_1_relu[0][0]']    
                                                                                                  
 conv3_block1_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_0_conv (Conv2D)   (None, 56, 56, 512)  131584      ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block1_2_relu[0][0]']    
                                                                                                  
 conv3_block1_0_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_0_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_0_bn[0][0]',      
                                                                  'conv3_block1_3_bn[0][0]']      
                                                                                                  
 conv3_block1_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block1_add[0][0]']       
                                                                                                  
 conv3_block2_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block1_out[0][0]']       
                                                                                                  
 conv3_block2_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block2_1_relu[0][0]']    
                                                                                                  
 conv3_block2_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block2_2_relu[0][0]']    
                                                                                                  
 conv3_block2_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block2_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_out[0][0]',       
                                                                  'conv3_block2_3_bn[0][0]']      
                                                                                                  
 conv3_block2_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block2_add[0][0]']       
                                                                                                  
 conv3_block3_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block2_out[0][0]']       
                                                                                                  
 conv3_block3_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block3_1_relu[0][0]']    
                                                                                                  
 conv3_block3_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block3_2_relu[0][0]']    
                                                                                                  
 conv3_block3_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block3_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_add (Add)         (None, 56, 56, 512)  0           ['conv3_block2_out[0][0]',       
                                                                  'conv3_block3_3_bn[0][0]']      
                                                                                                  
 conv3_block3_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block3_add[0][0]']       
                                                                                                  
 conv3_block4_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block3_out[0][0]']       
                                                                                                  
 conv3_block4_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block4_1_relu[0][0]']    
                                                                                                  
 conv3_block4_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block4_2_relu[0][0]']    
                                                                                                  
 conv3_block4_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block4_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_add (Add)         (None, 56, 56, 512)  0           ['conv3_block3_out[0][0]',       
                                                                  'conv3_block4_3_bn[0][0]']      
                                                                                                  
 conv3_block4_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block4_add[0][0]']       
                                                                                                  
 conv4_block1_1_conv (Conv2D)   (None, 28, 28, 256)  131328      ['conv3_block4_out[0][0]']       
                                                                                                  
 conv4_block1_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block1_1_relu[0][0]']    
                                                                                                  
 conv4_block1_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_0_conv (Conv2D)   (None, 28, 28, 1024  525312      ['conv3_block4_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block1_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block1_0_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_0_bn[0][0]',      
                                )                                 'conv4_block1_3_bn[0][0]']      
                                                                                                  
 conv4_block1_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block2_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block1_out[0][0]']       
                                                                                                  
 conv4_block2_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block2_1_relu[0][0]']    
                                                                                                  
 conv4_block2_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block2_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block2_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_out[0][0]',       
                                )                                 'conv4_block2_3_bn[0][0]']      
                                                                                                  
 conv4_block2_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block3_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block2_out[0][0]']       
                                                                                                  
 conv4_block3_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block3_1_relu[0][0]']    
                                                                                                  
 conv4_block3_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block3_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block3_add (Add)         (None, 28, 28, 1024  0           ['conv4_block2_out[0][0]',       
                                )                                 'conv4_block3_3_bn[0][0]']      
                                                                                                  
 conv4_block3_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block4_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block3_out[0][0]']       
                                                                                                  
 conv4_block4_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block4_1_relu[0][0]']    
                                                                                                  
 conv4_block4_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block4_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block4_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block4_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block4_add (Add)         (None, 28, 28, 1024  0           ['conv4_block3_out[0][0]',       
                                )                                 'conv4_block4_3_bn[0][0]']      
                                                                                                  
 conv4_block4_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block4_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block5_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block4_out[0][0]']       
                                                                                                  
 conv4_block5_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block5_1_relu[0][0]']    
                                                                                                  
 conv4_block5_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block5_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block5_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block5_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block5_add (Add)         (None, 28, 28, 1024  0           ['conv4_block4_out[0][0]',       
                                )                                 'conv4_block5_3_bn[0][0]']      
                                                                                                  
 conv4_block5_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block5_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block6_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block5_out[0][0]']       
                                                                                                  
 conv4_block6_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block6_1_relu[0][0]']    
                                                                                                  
 conv4_block6_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block6_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block6_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block6_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block6_add (Add)         (None, 28, 28, 1024  0           ['conv4_block5_out[0][0]',       
                                )                                 'conv4_block6_3_bn[0][0]']      
                                                                                                  
 conv4_block6_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block6_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_1_conv (Conv2D)   (None, 14, 14, 512)  524800      ['conv4_block6_out[0][0]']       
                                                                                                  
 conv5_block1_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block1_1_relu[0][0]']    
                                                                                                  
 conv5_block1_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_0_conv (Conv2D)   (None, 14, 14, 2048  2099200     ['conv4_block6_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block1_0_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_0_bn[0][0]',      
                                )                                 'conv5_block1_3_bn[0][0]']      
                                                                                                  
 conv5_block1_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block2_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block1_out[0][0]']       
                                                                                                  
 conv5_block2_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block2_1_relu[0][0]']    
                                                                                                  
 conv5_block2_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block2_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block2_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_out[0][0]',       
                                )                                 'conv5_block2_3_bn[0][0]']      
                                                                                                  
 conv5_block2_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block3_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block2_out[0][0]']       
                                                                                                  
 conv5_block3_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block3_1_relu[0][0]']    
                                                                                                  
 conv5_block3_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block3_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block3_add (Add)         (None, 14, 14, 2048  0           ['conv5_block2_out[0][0]',       
                                )                                 'conv5_block3_3_bn[0][0]']      
                                                                                                  
 conv5_block3_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 Flatten_for_hidden_layers (Glo  (None, 2048)        0           ['conv5_block3_out[0][0]']       
 balAveragePooling2D)                                                                             
                                                                                                  
 Dropout1 (Dropout)             (None, 2048)         0           ['Flatten_for_hidden_layers[0][0]
                                                                 ']                               
                                                                                                  
 Hidden_Layer1 (Dense)          (None, 128)          262272      ['Dropout1[0][0]']               
                                                                                                  
 Dropout2 (Dropout)             (None, 128)          0           ['Hidden_Layer1[0][0]']          
                                                                                                  
 Hidden_Layer2 (Dense)          (None, 128)          16512       ['Dropout2[0][0]']               
                                                                                                  
 output (Dense)                 (None, 1)            129         ['Hidden_Layer2[0][0]']          
                                                                                                  
==================================================================================================
Total params: 23,866,625
Trainable params: 283,009
Non-trainable params: 23,583,616
__________________________________________________________________________________________________


# plotting the model
plot_model(resnet_50, to_file='resnet_50.png', show_shapes=True, show_layer_names=True)


tf.autograph.experimental.do_not_convert(func=None)

<function tensorflow.python.autograph.impl.api.do_not_convert(func=None)>


# Size of TRAIN & VALIDATION labels and BATCH SIZE
y_train.shape[0], BATCH_SIZE, y_val.shape[0]

(53, 9, 10)


# Calculating train steps
train_steps = y_train.shape[0] // BATCH_SIZE
train_steps

5


# Calculating test steps
valid_steps = y_val.shape[0] // BATCH_SIZE
valid_steps

1


# Training the ResNet-50 model having custom top
history1 = resnet_50.fit(X_train, y_train,
                         epochs=15,
                         batch_size=BATCH_SIZE,
                         callbacks=[tensorboard_callback1, reduce_lr1],
                         steps_per_epoch=train_steps,
                         validation_steps=valid_steps,
                         validation_data=[X_val, y_val],
                         verbose=1)

Epoch 1/15
5/5 [==============================] - 49s 10s/step - loss: 7003.0762 - r_square: -29.9524 - val_loss: 4957.2769 - val_r_square: -25.1141 - lr: 1.0000e-04
Epoch 2/15
5/5 [==============================] - 38s 8s/step - loss: 6705.4595 - r_square: -21.9450 - val_loss: 4784.9590 - val_r_square: -24.2064 - lr: 1.0000e-04
Epoch 3/15
5/5 [==============================] - 48s 10s/step - loss: 6727.6392 - r_square: -26.8386 - val_loss: 4632.7388 - val_r_square: -23.4045 - lr: 1.0000e-04
Epoch 4/15
5/5 [==============================] - 47s 9s/step - loss: 6555.0513 - r_square: -27.5158 - val_loss: 4475.1504 - val_r_square: -22.5743 - lr: 1.0000e-04
Epoch 5/15
5/5 [==============================] - 39s 8s/step - loss: 6015.2046 - r_square: -27.5723 - val_loss: 4303.6104 - val_r_square: -21.6707 - lr: 1.0000e-04
Epoch 6/15
5/5 [==============================] - 44s 10s/step - loss: 6083.2095 - r_square: -24.5995 - val_loss: 4120.6885 - val_r_square: -20.7071 - lr: 1.0000e-04
Epoch 7/15
5/5 [==============================] - 45s 10s/step - loss: 5762.4517 - r_square: -21.2671 - val_loss: 3929.8545 - val_r_square: -19.7018 - lr: 1.0000e-04
Epoch 8/15
5/5 [==============================] - 47s 10s/step - loss: 5645.2061 - r_square: -19.9354 - val_loss: 3726.7668 - val_r_square: -18.6320 - lr: 1.0000e-04
Epoch 9/15
5/5 [==============================] - 44s 10s/step - loss: 5461.9282 - r_square: -20.5505 - val_loss: 3514.0061 - val_r_square: -17.5112 - lr: 1.0000e-04
Epoch 10/15
5/5 [==============================] - 46s 10s/step - loss: 4850.3354 - r_square: -19.0006 - val_loss: 3293.4675 - val_r_square: -16.3494 - lr: 1.0000e-04
Epoch 11/15
5/5 [==============================] - 44s 9s/step - loss: 4553.5239 - r_square: -18.1473 - val_loss: 3059.0615 - val_r_square: -15.1146 - lr: 1.0000e-04
Epoch 12/15
5/5 [==============================] - 44s 9s/step - loss: 4483.7969 - r_square: -16.0694 - val_loss: 2819.5720 - val_r_square: -13.8530 - lr: 1.0000e-04
Epoch 13/15
5/5 [==============================] - 45s 10s/step - loss: 4236.6973 - r_square: -14.5281 - val_loss: 2572.9409 - val_r_square: -12.5538 - lr: 1.0000e-04
Epoch 14/15
5/5 [==============================] - 39s 8s/step - loss: 3798.1116 - r_square: -15.6408 - val_loss: 2328.3533 - val_r_square: -11.2654 - lr: 1.0000e-04
Epoch 15/15
5/5 [==============================] - 43s 9s/step - loss: 3761.2983 - r_square: -18.9061 - val_loss: 2085.4873 - val_r_square: -9.9860 - lr: 1.0000e-04


# Storing the TRAIN & VAL Losses
train_losses = history1.history['loss']
validation_losses = history1.history['val_loss']


# Generate a sequence of integers to represent the epoch numbers
epochs = range(1, 16)

with plt.style.context('seaborn-poster'):
    plt.figure(figsize=(12,8))
    # Plot and label the training and validation loss values
    plt.plot(epochs, train_losses, label='Training Loss', marker='o', markersize=10)
    plt.plot(epochs, validation_losses, label='Validation Loss', marker='o', markersize=10)

    # Add in a title, grid and axes labels
    plt.title('Training & Validation Loss', fontdict=title_font_dict)
    plt.xlabel('Epochs', fontdict=label_font_dict)
    plt.ylabel('Mean Squared Loss', fontdict=label_font_dict)
    plt.grid(which='major', linestyle=":", color='lightpink')
    plt.minorticks_on()

    # Set the tick locations
    plt.xticks(arange(0, 16, 1))

    # Display the plot
    plt.legend(loc='best')
    plt.show()


# Storing the TRAIN & VAL R-square values
train_r2 = history1.history['r_square']
val_r2 = history1.history['val_r_square']

# Generate a sequence of integers to represent the epoch numbers
epochs = range(1, 16)

with plt.style.context('seaborn-poster'):
    plt.figure(figsize=(12,8))
    # Plot and label the training and validation r-square values
    plt.plot(epochs, train_r2, label='Training R2', marker='^', markersize=10)
    plt.plot(epochs, val_r2, label='Validation R2', marker='^', markersize=10)
    plt.grid(which='major', linestyle=":", color='lightpink')
    plt.minorticks_on()

    # Add in a title and axes labels
    plt.title('Training & Validation R-Square', fontdict=title_font_dict)
    plt.xlabel('Epochs', fontdict=label_font_dict)
    plt.ylabel('R-Square value', fontdict=label_font_dict)

    # Set the tick locations
    plt.xticks(arange(0, 16, 1))

    # Display the plot
    plt.legend(loc='best')
    plt.show()


# Generating the model predictions on the validation set
val_heights = resnet_50.predict(X_val)
val_preds = []
for val in val_heights:
    val_preds.append((val[0]))

val_preds = np.array(val_preds)
val_preds

1/1 [==============================] - 8s 8s/step

array([26.534063, 26.688957, 26.343487, 26.851963, 26.60034 , 26.936205,
       26.607258, 26.310953, 26.984158, 26.613827], dtype=float32)


# Actual VAL Tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# Validation R2 Score
r2_score(y_val, val_preds)

-10.469640721864407


# Generating the model predictions on the TEST set
test_heights = resnet_50.predict(X_test)
test_preds = []
for val in test_heights:
    test_preds.append((val[0]))

test_preds = np.array(test_preds)
test_preds

1/1 [==============================] - 1s 1s/step

array([26.642708, 26.367102], dtype=float32)


# Actual TEST Tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# TEST R2 Score
r2_score(y_test, test_preds)

-99.50112793638546


# Fetching the directory name
curr_run_logdir1.split("/")[-1]

'run_2022_12_04-06_53_08'


# Any running instances
notebook.list()

No known TensorBoard instances running.


# Run TF Board
%tensorboard --logdir logs


# Loading the required libraries
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import mean_squared_error, mean_absolute_error, mean_absolute_percentage_error
from sklearn.ensemble import RandomForestRegressor


# build the ResNet-50 network
resnet50_with_no_top_model = tf.keras.applications.ResNet50(include_top=False, weights='imagenet', input_shape=(448,448,3))


# Model summary
resnet50_with_no_top_model.summary()

Model: "resnet50"
__________________________________________________________________________________________________
 Layer (type)                   Output Shape         Param #     Connected to                     
==================================================================================================
 input_1 (InputLayer)           [(None, 448, 448, 3  0           []                               
                                )]                                                                
                                                                                                  
 conv1_pad (ZeroPadding2D)      (None, 454, 454, 3)  0           ['input_1[0][0]']                
                                                                                                  
 conv1_conv (Conv2D)            (None, 224, 224, 64  9472        ['conv1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv1_bn (BatchNormalization)  (None, 224, 224, 64  256         ['conv1_conv[0][0]']             
                                )                                                                 
                                                                                                  
 conv1_relu (Activation)        (None, 224, 224, 64  0           ['conv1_bn[0][0]']               
                                )                                                                 
                                                                                                  
 pool1_pad (ZeroPadding2D)      (None, 226, 226, 64  0           ['conv1_relu[0][0]']             
                                )                                                                 
                                                                                                  
 pool1_pool (MaxPooling2D)      (None, 112, 112, 64  0           ['pool1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv2_block1_1_conv (Conv2D)   (None, 112, 112, 64  4160        ['pool1_pool[0][0]']             
                                )                                                                 
                                                                                                  
 conv2_block1_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block1_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block1_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_0_conv (Conv2D)   (None, 112, 112, 25  16640       ['pool1_pool[0][0]']             
                                6)                                                                
                                                                                                  
 conv2_block1_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block1_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block1_0_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_0_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_0_bn[0][0]',      
                                6)                                'conv2_block1_3_bn[0][0]']      
                                                                                                  
 conv2_block1_out (Activation)  (None, 112, 112, 25  0           ['conv2_block1_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block2_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block1_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block2_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block2_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block2_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block2_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block2_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block2_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block2_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_out[0][0]',       
                                6)                                'conv2_block2_3_bn[0][0]']      
                                                                                                  
 conv2_block2_out (Activation)  (None, 112, 112, 25  0           ['conv2_block2_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block3_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block2_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block3_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block3_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block3_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block3_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block3_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block3_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block3_add (Add)         (None, 112, 112, 25  0           ['conv2_block2_out[0][0]',       
                                6)                                'conv2_block3_3_bn[0][0]']      
                                                                                                  
 conv2_block3_out (Activation)  (None, 112, 112, 25  0           ['conv2_block3_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv3_block1_1_conv (Conv2D)   (None, 56, 56, 128)  32896       ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block1_1_relu[0][0]']    
                                                                                                  
 conv3_block1_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_0_conv (Conv2D)   (None, 56, 56, 512)  131584      ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block1_2_relu[0][0]']    
                                                                                                  
 conv3_block1_0_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_0_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_0_bn[0][0]',      
                                                                  'conv3_block1_3_bn[0][0]']      
                                                                                                  
 conv3_block1_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block1_add[0][0]']       
                                                                                                  
 conv3_block2_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block1_out[0][0]']       
                                                                                                  
 conv3_block2_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block2_1_relu[0][0]']    
                                                                                                  
 conv3_block2_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block2_2_relu[0][0]']    
                                                                                                  
 conv3_block2_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block2_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_out[0][0]',       
                                                                  'conv3_block2_3_bn[0][0]']      
                                                                                                  
 conv3_block2_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block2_add[0][0]']       
                                                                                                  
 conv3_block3_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block2_out[0][0]']       
                                                                                                  
 conv3_block3_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block3_1_relu[0][0]']    
                                                                                                  
 conv3_block3_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block3_2_relu[0][0]']    
                                                                                                  
 conv3_block3_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block3_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_add (Add)         (None, 56, 56, 512)  0           ['conv3_block2_out[0][0]',       
                                                                  'conv3_block3_3_bn[0][0]']      
                                                                                                  
 conv3_block3_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block3_add[0][0]']       
                                                                                                  
 conv3_block4_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block3_out[0][0]']       
                                                                                                  
 conv3_block4_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block4_1_relu[0][0]']    
                                                                                                  
 conv3_block4_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block4_2_relu[0][0]']    
                                                                                                  
 conv3_block4_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block4_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_add (Add)         (None, 56, 56, 512)  0           ['conv3_block3_out[0][0]',       
                                                                  'conv3_block4_3_bn[0][0]']      
                                                                                                  
 conv3_block4_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block4_add[0][0]']       
                                                                                                  
 conv4_block1_1_conv (Conv2D)   (None, 28, 28, 256)  131328      ['conv3_block4_out[0][0]']       
                                                                                                  
 conv4_block1_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block1_1_relu[0][0]']    
                                                                                                  
 conv4_block1_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_0_conv (Conv2D)   (None, 28, 28, 1024  525312      ['conv3_block4_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block1_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block1_0_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_0_bn[0][0]',      
                                )                                 'conv4_block1_3_bn[0][0]']      
                                                                                                  
 conv4_block1_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block2_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block1_out[0][0]']       
                                                                                                  
 conv4_block2_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block2_1_relu[0][0]']    
                                                                                                  
 conv4_block2_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block2_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block2_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_out[0][0]',       
                                )                                 'conv4_block2_3_bn[0][0]']      
                                                                                                  
 conv4_block2_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block3_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block2_out[0][0]']       
                                                                                                  
 conv4_block3_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block3_1_relu[0][0]']    
                                                                                                  
 conv4_block3_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block3_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block3_add (Add)         (None, 28, 28, 1024  0           ['conv4_block2_out[0][0]',       
                                )                                 'conv4_block3_3_bn[0][0]']      
                                                                                                  
 conv4_block3_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block4_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block3_out[0][0]']       
                                                                                                  
 conv4_block4_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block4_1_relu[0][0]']    
                                                                                                  
 conv4_block4_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block4_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block4_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block4_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block4_add (Add)         (None, 28, 28, 1024  0           ['conv4_block3_out[0][0]',       
                                )                                 'conv4_block4_3_bn[0][0]']      
                                                                                                  
 conv4_block4_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block4_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block5_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block4_out[0][0]']       
                                                                                                  
 conv4_block5_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block5_1_relu[0][0]']    
                                                                                                  
 conv4_block5_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block5_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block5_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block5_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block5_add (Add)         (None, 28, 28, 1024  0           ['conv4_block4_out[0][0]',       
                                )                                 'conv4_block5_3_bn[0][0]']      
                                                                                                  
 conv4_block5_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block5_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block6_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block5_out[0][0]']       
                                                                                                  
 conv4_block6_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block6_1_relu[0][0]']    
                                                                                                  
 conv4_block6_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block6_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block6_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block6_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block6_add (Add)         (None, 28, 28, 1024  0           ['conv4_block5_out[0][0]',       
                                )                                 'conv4_block6_3_bn[0][0]']      
                                                                                                  
 conv4_block6_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block6_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_1_conv (Conv2D)   (None, 14, 14, 512)  524800      ['conv4_block6_out[0][0]']       
                                                                                                  
 conv5_block1_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block1_1_relu[0][0]']    
                                                                                                  
 conv5_block1_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_0_conv (Conv2D)   (None, 14, 14, 2048  2099200     ['conv4_block6_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block1_0_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_0_bn[0][0]',      
                                )                                 'conv5_block1_3_bn[0][0]']      
                                                                                                  
 conv5_block1_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block2_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block1_out[0][0]']       
                                                                                                  
 conv5_block2_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block2_1_relu[0][0]']    
                                                                                                  
 conv5_block2_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block2_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block2_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_out[0][0]',       
                                )                                 'conv5_block2_3_bn[0][0]']      
                                                                                                  
 conv5_block2_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block3_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block2_out[0][0]']       
                                                                                                  
 conv5_block3_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block3_1_relu[0][0]']    
                                                                                                  
 conv5_block3_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block3_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block3_add (Add)         (None, 14, 14, 2048  0           ['conv5_block2_out[0][0]',       
                                )                                 'conv5_block3_3_bn[0][0]']      
                                                                                                  
 conv5_block3_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
==================================================================================================
Total params: 23,587,712
Trainable params: 23,534,592
Non-trainable params: 53,120
__________________________________________________________________________________________________


from tensorflow.keras.callbacks import ModelCheckpoint


# Instantiating Optimizer
learning_rate= 0.0001
opt2 = tf.keras.optimizers.Adam(learning_rate=learning_rate)

# Reduce Learning Rate on Plateau
reduce_lr2 = tf.keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.1, patience=2, verbose=1, mode='auto', min_delta=0.0001)

# Checkpointing
filepath = "weights-{epoch:02d}-{val_loss:.2f}.hdf5"
checkpoint = ModelCheckpoint(filepath, monitor='val_loss', verbose=1, save_best_only=True, mode='auto')


# Logs directory
curr_run_logdir2 = get_run_logdir()

# Instantiating Tensorboard callback
tensorboard_callback2 = TensorBoard(log_dir=curr_run_logdir2, histogram_freq=1)


# Setting the seed
os.environ['PYTHONHASHSEED'] = '0'

# Clearing the TF session
tf.keras.backend.clear_session()

# defining the custom top of the ResNet-50 model
for layer in resnet50_with_no_top_model.layers:
    layer.trainable = False

# Adding additional layers
input_layer = resnet50_with_no_top_model.output
    
# Defining the top layers structure of the model
flatten = tf.keras.layers.GlobalAveragePooling2D(name='Flatten_for_hidden_layers')(input_layer)

dropout_1 = Dropout(rate=0.5, name='Dropout1')(flatten)

dense_layer1 = tf.keras.layers.Dense(units=64, 
                                     activation='relu', 
                                     use_bias=True, 
                                     kernel_initializer=tf.keras.initializers.he_normal(seed=80),
                                     bias_initializer=tf.keras.initializers.he_normal(seed=110),
                                     name='Hidden_Layer1')(dropout_1)

dropout_2 = Dropout(rate=0.5, name='Dropout2')(dense_layer1)

dense_layer2 = tf.keras.layers.Dense(units=128, 
                                     activation='relu', 
                                     use_bias=True, 
                                     kernel_initializer=tf.keras.initializers.he_normal(seed=80),
                                     bias_initializer=tf.keras.initializers.he_normal(seed=110),
                                     name='Hidden_Layer2')(dropout_2)

output_layer = tf.keras.layers.Dense(1, activation='linear', name="output")(dense_layer2)

# Instantiating the complete model
resnet_50b = Model(inputs=resnet50_with_no_top_model.input, outputs=output_layer)

# Compiling the model
resnet_50b.compile(optimizer=opt2,
                  loss = 'mean_squared_error',
                  metrics=[r_square])

# Summary of the ResNet-50 model with custom top
resnet_50b.summary()

Model: "model"
__________________________________________________________________________________________________
 Layer (type)                   Output Shape         Param #     Connected to                     
==================================================================================================
 input_1 (InputLayer)           [(None, 448, 448, 3  0           []                               
                                )]                                                                
                                                                                                  
 conv1_pad (ZeroPadding2D)      (None, 454, 454, 3)  0           ['input_1[0][0]']                
                                                                                                  
 conv1_conv (Conv2D)            (None, 224, 224, 64  9472        ['conv1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv1_bn (BatchNormalization)  (None, 224, 224, 64  256         ['conv1_conv[0][0]']             
                                )                                                                 
                                                                                                  
 conv1_relu (Activation)        (None, 224, 224, 64  0           ['conv1_bn[0][0]']               
                                )                                                                 
                                                                                                  
 pool1_pad (ZeroPadding2D)      (None, 226, 226, 64  0           ['conv1_relu[0][0]']             
                                )                                                                 
                                                                                                  
 pool1_pool (MaxPooling2D)      (None, 112, 112, 64  0           ['pool1_pad[0][0]']              
                                )                                                                 
                                                                                                  
 conv2_block1_1_conv (Conv2D)   (None, 112, 112, 64  4160        ['pool1_pool[0][0]']             
                                )                                                                 
                                                                                                  
 conv2_block1_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block1_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block1_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block1_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block1_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block1_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block1_0_conv (Conv2D)   (None, 112, 112, 25  16640       ['pool1_pool[0][0]']             
                                6)                                                                
                                                                                                  
 conv2_block1_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block1_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block1_0_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_0_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block1_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block1_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_0_bn[0][0]',      
                                6)                                'conv2_block1_3_bn[0][0]']      
                                                                                                  
 conv2_block1_out (Activation)  (None, 112, 112, 25  0           ['conv2_block1_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block2_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block1_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block2_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block2_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block2_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block2_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block2_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block2_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block2_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block2_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block2_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block2_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block2_add (Add)         (None, 112, 112, 25  0           ['conv2_block1_out[0][0]',       
                                6)                                'conv2_block2_3_bn[0][0]']      
                                                                                                  
 conv2_block2_out (Activation)  (None, 112, 112, 25  0           ['conv2_block2_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv2_block3_1_conv (Conv2D)   (None, 112, 112, 64  16448       ['conv2_block2_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv2_block3_1_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_1_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_1_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_1_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_2_conv (Conv2D)   (None, 112, 112, 64  36928       ['conv2_block3_1_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv2_block3_2_bn (BatchNormal  (None, 112, 112, 64  256        ['conv2_block3_2_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv2_block3_2_relu (Activatio  (None, 112, 112, 64  0          ['conv2_block3_2_bn[0][0]']      
 n)                             )                                                                 
                                                                                                  
 conv2_block3_3_conv (Conv2D)   (None, 112, 112, 25  16640       ['conv2_block3_2_relu[0][0]']    
                                6)                                                                
                                                                                                  
 conv2_block3_3_bn (BatchNormal  (None, 112, 112, 25  1024       ['conv2_block3_3_conv[0][0]']    
 ization)                       6)                                                                
                                                                                                  
 conv2_block3_add (Add)         (None, 112, 112, 25  0           ['conv2_block2_out[0][0]',       
                                6)                                'conv2_block3_3_bn[0][0]']      
                                                                                                  
 conv2_block3_out (Activation)  (None, 112, 112, 25  0           ['conv2_block3_add[0][0]']       
                                6)                                                                
                                                                                                  
 conv3_block1_1_conv (Conv2D)   (None, 56, 56, 128)  32896       ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block1_1_relu[0][0]']    
                                                                                                  
 conv3_block1_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block1_0_conv (Conv2D)   (None, 56, 56, 512)  131584      ['conv2_block3_out[0][0]']       
                                                                                                  
 conv3_block1_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block1_2_relu[0][0]']    
                                                                                                  
 conv3_block1_0_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_0_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block1_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block1_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_0_bn[0][0]',      
                                                                  'conv3_block1_3_bn[0][0]']      
                                                                                                  
 conv3_block1_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block1_add[0][0]']       
                                                                                                  
 conv3_block2_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block1_out[0][0]']       
                                                                                                  
 conv3_block2_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block2_1_relu[0][0]']    
                                                                                                  
 conv3_block2_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block2_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block2_2_relu[0][0]']    
                                                                                                  
 conv3_block2_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block2_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block2_add (Add)         (None, 56, 56, 512)  0           ['conv3_block1_out[0][0]',       
                                                                  'conv3_block2_3_bn[0][0]']      
                                                                                                  
 conv3_block2_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block2_add[0][0]']       
                                                                                                  
 conv3_block3_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block2_out[0][0]']       
                                                                                                  
 conv3_block3_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block3_1_relu[0][0]']    
                                                                                                  
 conv3_block3_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block3_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block3_2_relu[0][0]']    
                                                                                                  
 conv3_block3_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block3_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block3_add (Add)         (None, 56, 56, 512)  0           ['conv3_block2_out[0][0]',       
                                                                  'conv3_block3_3_bn[0][0]']      
                                                                                                  
 conv3_block3_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block3_add[0][0]']       
                                                                                                  
 conv3_block4_1_conv (Conv2D)   (None, 56, 56, 128)  65664       ['conv3_block3_out[0][0]']       
                                                                                                  
 conv3_block4_1_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_1_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_2_conv (Conv2D)   (None, 56, 56, 128)  147584      ['conv3_block4_1_relu[0][0]']    
                                                                                                  
 conv3_block4_2_bn (BatchNormal  (None, 56, 56, 128)  512        ['conv3_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_2_relu (Activatio  (None, 56, 56, 128)  0          ['conv3_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv3_block4_3_conv (Conv2D)   (None, 56, 56, 512)  66048       ['conv3_block4_2_relu[0][0]']    
                                                                                                  
 conv3_block4_3_bn (BatchNormal  (None, 56, 56, 512)  2048       ['conv3_block4_3_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv3_block4_add (Add)         (None, 56, 56, 512)  0           ['conv3_block3_out[0][0]',       
                                                                  'conv3_block4_3_bn[0][0]']      
                                                                                                  
 conv3_block4_out (Activation)  (None, 56, 56, 512)  0           ['conv3_block4_add[0][0]']       
                                                                                                  
 conv4_block1_1_conv (Conv2D)   (None, 28, 28, 256)  131328      ['conv3_block4_out[0][0]']       
                                                                                                  
 conv4_block1_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block1_1_relu[0][0]']    
                                                                                                  
 conv4_block1_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block1_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block1_0_conv (Conv2D)   (None, 28, 28, 1024  525312      ['conv3_block4_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block1_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block1_0_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block1_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_0_bn[0][0]',      
                                )                                 'conv4_block1_3_bn[0][0]']      
                                                                                                  
 conv4_block1_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block2_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block1_out[0][0]']       
                                                                                                  
 conv4_block2_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block2_1_relu[0][0]']    
                                                                                                  
 conv4_block2_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block2_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block2_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block2_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block2_add (Add)         (None, 28, 28, 1024  0           ['conv4_block1_out[0][0]',       
                                )                                 'conv4_block2_3_bn[0][0]']      
                                                                                                  
 conv4_block2_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block3_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block2_out[0][0]']       
                                                                                                  
 conv4_block3_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block3_1_relu[0][0]']    
                                                                                                  
 conv4_block3_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block3_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block3_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block3_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block3_add (Add)         (None, 28, 28, 1024  0           ['conv4_block2_out[0][0]',       
                                )                                 'conv4_block3_3_bn[0][0]']      
                                                                                                  
 conv4_block3_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block4_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block3_out[0][0]']       
                                                                                                  
 conv4_block4_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block4_1_relu[0][0]']    
                                                                                                  
 conv4_block4_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block4_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block4_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block4_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block4_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block4_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block4_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block4_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block4_add (Add)         (None, 28, 28, 1024  0           ['conv4_block3_out[0][0]',       
                                )                                 'conv4_block4_3_bn[0][0]']      
                                                                                                  
 conv4_block4_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block4_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block5_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block4_out[0][0]']       
                                                                                                  
 conv4_block5_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block5_1_relu[0][0]']    
                                                                                                  
 conv4_block5_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block5_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block5_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block5_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block5_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block5_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block5_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block5_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block5_add (Add)         (None, 28, 28, 1024  0           ['conv4_block4_out[0][0]',       
                                )                                 'conv4_block5_3_bn[0][0]']      
                                                                                                  
 conv4_block5_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block5_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv4_block6_1_conv (Conv2D)   (None, 28, 28, 256)  262400      ['conv4_block5_out[0][0]']       
                                                                                                  
 conv4_block6_1_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_1_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_2_conv (Conv2D)   (None, 28, 28, 256)  590080      ['conv4_block6_1_relu[0][0]']    
                                                                                                  
 conv4_block6_2_bn (BatchNormal  (None, 28, 28, 256)  1024       ['conv4_block6_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv4_block6_2_relu (Activatio  (None, 28, 28, 256)  0          ['conv4_block6_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv4_block6_3_conv (Conv2D)   (None, 28, 28, 1024  263168      ['conv4_block6_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv4_block6_3_bn (BatchNormal  (None, 28, 28, 1024  4096       ['conv4_block6_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv4_block6_add (Add)         (None, 28, 28, 1024  0           ['conv4_block5_out[0][0]',       
                                )                                 'conv4_block6_3_bn[0][0]']      
                                                                                                  
 conv4_block6_out (Activation)  (None, 28, 28, 1024  0           ['conv4_block6_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_1_conv (Conv2D)   (None, 14, 14, 512)  524800      ['conv4_block6_out[0][0]']       
                                                                                                  
 conv5_block1_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block1_1_relu[0][0]']    
                                                                                                  
 conv5_block1_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block1_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block1_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block1_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block1_0_conv (Conv2D)   (None, 14, 14, 2048  2099200     ['conv4_block6_out[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block1_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block1_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block1_0_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_0_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block1_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block1_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_0_bn[0][0]',      
                                )                                 'conv5_block1_3_bn[0][0]']      
                                                                                                  
 conv5_block1_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block1_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block2_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block1_out[0][0]']       
                                                                                                  
 conv5_block2_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block2_1_relu[0][0]']    
                                                                                                  
 conv5_block2_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block2_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block2_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block2_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block2_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block2_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block2_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block2_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block2_add (Add)         (None, 14, 14, 2048  0           ['conv5_block1_out[0][0]',       
                                )                                 'conv5_block2_3_bn[0][0]']      
                                                                                                  
 conv5_block2_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block2_add[0][0]']       
                                )                                                                 
                                                                                                  
 conv5_block3_1_conv (Conv2D)   (None, 14, 14, 512)  1049088     ['conv5_block2_out[0][0]']       
                                                                                                  
 conv5_block3_1_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_1_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_1_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_1_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_2_conv (Conv2D)   (None, 14, 14, 512)  2359808     ['conv5_block3_1_relu[0][0]']    
                                                                                                  
 conv5_block3_2_bn (BatchNormal  (None, 14, 14, 512)  2048       ['conv5_block3_2_conv[0][0]']    
 ization)                                                                                         
                                                                                                  
 conv5_block3_2_relu (Activatio  (None, 14, 14, 512)  0          ['conv5_block3_2_bn[0][0]']      
 n)                                                                                               
                                                                                                  
 conv5_block3_3_conv (Conv2D)   (None, 14, 14, 2048  1050624     ['conv5_block3_2_relu[0][0]']    
                                )                                                                 
                                                                                                  
 conv5_block3_3_bn (BatchNormal  (None, 14, 14, 2048  8192       ['conv5_block3_3_conv[0][0]']    
 ization)                       )                                                                 
                                                                                                  
 conv5_block3_add (Add)         (None, 14, 14, 2048  0           ['conv5_block2_out[0][0]',       
                                )                                 'conv5_block3_3_bn[0][0]']      
                                                                                                  
 conv5_block3_out (Activation)  (None, 14, 14, 2048  0           ['conv5_block3_add[0][0]']       
                                )                                                                 
                                                                                                  
 Flatten_for_hidden_layers (Glo  (None, 2048)        0           ['conv5_block3_out[0][0]']       
 balAveragePooling2D)                                                                             
                                                                                                  
 Dropout1 (Dropout)             (None, 2048)         0           ['Flatten_for_hidden_layers[0][0]
                                                                 ']                               
                                                                                                  
 Hidden_Layer1 (Dense)          (None, 64)           131136      ['Dropout1[0][0]']               
                                                                                                  
 Dropout2 (Dropout)             (None, 64)           0           ['Hidden_Layer1[0][0]']          
                                                                                                  
 Hidden_Layer2 (Dense)          (None, 128)          8320        ['Dropout2[0][0]']               
                                                                                                  
 output (Dense)                 (None, 1)            129         ['Hidden_Layer2[0][0]']          
                                                                                                  
==================================================================================================
Total params: 23,727,297
Trainable params: 139,585
Non-trainable params: 23,587,712
__________________________________________________________________________________________________


# serialize model to JSON
model_json = resnet_50b.to_json()


# Saving the resnet 50 architecture as a JSON file
import json
with open('resnet_50b.json', 'w') as json_file:
    json_file.write(model_json)


# plotting the model
plot_model(resnet_50b, to_file='resnet_50b.png', show_shapes=True, show_layer_names=True)


tf.autograph.experimental.do_not_convert(func=None)

<function tensorflow.python.autograph.impl.api.do_not_convert(func=None)>


# Size of TRAIN & VALIDATION labels and BATCH SIZE
y_train.shape[0], BATCH_SIZE, y_val.shape[0]

(53, 9, 10)


# Calculating train steps
train_steps = y_train.shape[0] // BATCH_SIZE
train_steps

5


# Calculating test steps
valid_steps = y_val.shape[0] // BATCH_SIZE
valid_steps

1


# Training the ResNet-50 model having custom top
history2 = resnet_50b.fit(X_train, y_train,
                         epochs=30,
                         batch_size=BATCH_SIZE,
                         callbacks=[tensorboard_callback2, reduce_lr2, checkpoint],
                         steps_per_epoch=train_steps,
                         validation_steps=valid_steps,
                         validation_data=[X_val, y_val],
                         verbose=1)

Epoch 1/30
5/5 [==============================] - ETA: 0s - loss: 6198.6245 - r_square: -22.7924
Epoch 1: val_loss improved from inf to 4428.71338, saving model to weights-01-4428.71.hdf5
5/5 [==============================] - 18s 2s/step - loss: 6198.6245 - r_square: -22.7924 - val_loss: 4428.7134 - val_r_square: -22.3297 - lr: 1.0000e-04
Epoch 2/30
5/5 [==============================] - ETA: 0s - loss: 6084.4565 - r_square: -22.1090
Epoch 2: val_loss improved from 4428.71338 to 4308.65234, saving model to weights-02-4308.65.hdf5
5/5 [==============================] - 7s 1s/step - loss: 6084.4565 - r_square: -22.1090 - val_loss: 4308.6523 - val_r_square: -21.6973 - lr: 1.0000e-04
Epoch 3/30
5/5 [==============================] - ETA: 0s - loss: 6203.9785 - r_square: -21.3119
Epoch 3: val_loss improved from 4308.65234 to 4185.42871, saving model to weights-03-4185.43.hdf5
5/5 [==============================] - 6s 1s/step - loss: 6203.9785 - r_square: -21.3119 - val_loss: 4185.4287 - val_r_square: -21.0481 - lr: 1.0000e-04
Epoch 4/30
5/5 [==============================] - ETA: 0s - loss: 5134.3301 - r_square: -20.0523
Epoch 4: val_loss improved from 4185.42871 to 4062.46191, saving model to weights-04-4062.46.hdf5
5/5 [==============================] - 6s 2s/step - loss: 5134.3301 - r_square: -20.0523 - val_loss: 4062.4619 - val_r_square: -20.4004 - lr: 1.0000e-04
Epoch 5/30
5/5 [==============================] - ETA: 0s - loss: 5462.3887 - r_square: -19.3322
Epoch 5: val_loss improved from 4062.46191 to 3922.05029, saving model to weights-05-3922.05.hdf5
5/5 [==============================] - 6s 1s/step - loss: 5462.3887 - r_square: -19.3322 - val_loss: 3922.0503 - val_r_square: -19.6607 - lr: 1.0000e-04
Epoch 6/30
5/5 [==============================] - ETA: 0s - loss: 5241.5405 - r_square: -20.6259
Epoch 6: val_loss improved from 3922.05029 to 3773.99121, saving model to weights-06-3773.99.hdf5
5/5 [==============================] - 6s 1s/step - loss: 5241.5405 - r_square: -20.6259 - val_loss: 3773.9912 - val_r_square: -18.8808 - lr: 1.0000e-04
Epoch 7/30
5/5 [==============================] - ETA: 0s - loss: 5181.4824 - r_square: -30.2086
Epoch 7: val_loss improved from 3773.99121 to 3612.79761, saving model to weights-07-3612.80.hdf5
5/5 [==============================] - 6s 1s/step - loss: 5181.4824 - r_square: -30.2086 - val_loss: 3612.7976 - val_r_square: -18.0316 - lr: 1.0000e-04
Epoch 8/30
5/5 [==============================] - ETA: 0s - loss: 4815.0923 - r_square: -18.8923
Epoch 8: val_loss improved from 3612.79761 to 3439.90894, saving model to weights-08-3439.91.hdf5
5/5 [==============================] - 6s 1s/step - loss: 4815.0923 - r_square: -18.8923 - val_loss: 3439.9089 - val_r_square: -17.1209 - lr: 1.0000e-04
Epoch 9/30
5/5 [==============================] - ETA: 0s - loss: 4559.5293 - r_square: -14.0418
Epoch 9: val_loss improved from 3439.90894 to 3247.51929, saving model to weights-09-3247.52.hdf5
5/5 [==============================] - 6s 1s/step - loss: 4559.5293 - r_square: -14.0418 - val_loss: 3247.5193 - val_r_square: -16.1074 - lr: 1.0000e-04
Epoch 10/30
5/5 [==============================] - ETA: 0s - loss: 4029.8274 - r_square: -13.3927
Epoch 10: val_loss improved from 3247.51929 to 3049.11133, saving model to weights-10-3049.11.hdf5
5/5 [==============================] - 6s 1s/step - loss: 4029.8274 - r_square: -13.3927 - val_loss: 3049.1113 - val_r_square: -15.0622 - lr: 1.0000e-04
Epoch 11/30
5/5 [==============================] - ETA: 0s - loss: 4219.2373 - r_square: -17.6608
Epoch 11: val_loss improved from 3049.11133 to 2845.84473, saving model to weights-11-2845.84.hdf5
5/5 [==============================] - 6s 1s/step - loss: 4219.2373 - r_square: -17.6608 - val_loss: 2845.8447 - val_r_square: -13.9914 - lr: 1.0000e-04
Epoch 12/30
5/5 [==============================] - ETA: 0s - loss: 4103.9746 - r_square: -25.7640
Epoch 12: val_loss improved from 2845.84473 to 2636.79980, saving model to weights-12-2636.80.hdf5
5/5 [==============================] - 6s 1s/step - loss: 4103.9746 - r_square: -25.7640 - val_loss: 2636.7998 - val_r_square: -12.8902 - lr: 1.0000e-04
Epoch 13/30
5/5 [==============================] - ETA: 0s - loss: 3448.9885 - r_square: -12.8468
Epoch 13: val_loss improved from 2636.79980 to 2420.93237, saving model to weights-13-2420.93.hdf5
5/5 [==============================] - 6s 1s/step - loss: 3448.9885 - r_square: -12.8468 - val_loss: 2420.9324 - val_r_square: -11.7531 - lr: 1.0000e-04
Epoch 14/30
5/5 [==============================] - ETA: 0s - loss: 3695.1414 - r_square: -21.5406
Epoch 14: val_loss improved from 2420.93237 to 2207.13403, saving model to weights-14-2207.13.hdf5
5/5 [==============================] - 6s 1s/step - loss: 3695.1414 - r_square: -21.5406 - val_loss: 2207.1340 - val_r_square: -10.6268 - lr: 1.0000e-04
Epoch 15/30
5/5 [==============================] - ETA: 0s - loss: 3130.0139 - r_square: -10.0164
Epoch 15: val_loss improved from 2207.13403 to 1997.55164, saving model to weights-15-1997.55.hdf5
5/5 [==============================] - 6s 1s/step - loss: 3130.0139 - r_square: -10.0164 - val_loss: 1997.5516 - val_r_square: -9.5228 - lr: 1.0000e-04
Epoch 16/30
5/5 [==============================] - ETA: 0s - loss: 2940.1853 - r_square: -10.3711
Epoch 16: val_loss improved from 1997.55164 to 1791.51172, saving model to weights-16-1791.51.hdf5
5/5 [==============================] - 6s 1s/step - loss: 2940.1853 - r_square: -10.3711 - val_loss: 1791.5117 - val_r_square: -8.4374 - lr: 1.0000e-04
Epoch 17/30
5/5 [==============================] - ETA: 0s - loss: 2543.3757 - r_square: -8.3292
Epoch 17: val_loss improved from 1791.51172 to 1589.50781, saving model to weights-17-1589.51.hdf5
5/5 [==============================] - 6s 1s/step - loss: 2543.3757 - r_square: -8.3292 - val_loss: 1589.5078 - val_r_square: -7.3733 - lr: 1.0000e-04
Epoch 18/30
5/5 [==============================] - ETA: 0s - loss: 2391.4343 - r_square: -8.4035
Epoch 18: val_loss improved from 1589.50781 to 1391.94861, saving model to weights-18-1391.95.hdf5
5/5 [==============================] - 6s 1s/step - loss: 2391.4343 - r_square: -8.4035 - val_loss: 1391.9486 - val_r_square: -6.3326 - lr: 1.0000e-04
Epoch 19/30
5/5 [==============================] - ETA: 0s - loss: 2088.0647 - r_square: -6.9122
Epoch 19: val_loss improved from 1391.94861 to 1203.21533, saving model to weights-19-1203.22.hdf5
5/5 [==============================] - 6s 1s/step - loss: 2088.0647 - r_square: -6.9122 - val_loss: 1203.2153 - val_r_square: -5.3383 - lr: 1.0000e-04
Epoch 20/30
5/5 [==============================] - ETA: 0s - loss: 2021.9028 - r_square: -6.7778
Epoch 20: val_loss improved from 1203.21533 to 1027.32593, saving model to weights-20-1027.33.hdf5
5/5 [==============================] - 6s 1s/step - loss: 2021.9028 - r_square: -6.7778 - val_loss: 1027.3259 - val_r_square: -4.4118 - lr: 1.0000e-04
Epoch 21/30
5/5 [==============================] - ETA: 0s - loss: 1685.9087 - r_square: -5.8522
Epoch 21: val_loss improved from 1027.32593 to 867.36707, saving model to weights-21-867.37.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1685.9087 - r_square: -5.8522 - val_loss: 867.3671 - val_r_square: -3.5691 - lr: 1.0000e-04
Epoch 22/30
5/5 [==============================] - ETA: 0s - loss: 1519.0421 - r_square: -5.0876
Epoch 22: val_loss improved from 867.36707 to 720.95288, saving model to weights-22-720.95.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1519.0421 - r_square: -5.0876 - val_loss: 720.9529 - val_r_square: -2.7979 - lr: 1.0000e-04
Epoch 23/30
5/5 [==============================] - ETA: 0s - loss: 1336.3978 - r_square: -3.7004
Epoch 23: val_loss improved from 720.95288 to 591.93451, saving model to weights-23-591.93.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1336.3978 - r_square: -3.7004 - val_loss: 591.9345 - val_r_square: -2.1182 - lr: 1.0000e-04
Epoch 24/30
5/5 [==============================] - ETA: 0s - loss: 1247.7491 - r_square: -3.9636
Epoch 24: val_loss improved from 591.93451 to 482.02951, saving model to weights-24-482.03.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1247.7491 - r_square: -3.9636 - val_loss: 482.0295 - val_r_square: -1.5393 - lr: 1.0000e-04
Epoch 25/30
5/5 [==============================] - ETA: 0s - loss: 1174.1055 - r_square: -3.4866
Epoch 25: val_loss improved from 482.02951 to 389.63132, saving model to weights-25-389.63.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1174.1055 - r_square: -3.4866 - val_loss: 389.6313 - val_r_square: -1.0525 - lr: 1.0000e-04
Epoch 26/30
5/5 [==============================] - ETA: 0s - loss: 851.1902 - r_square: -2.6707
Epoch 26: val_loss improved from 389.63132 to 318.95020, saving model to weights-26-318.95.hdf5
5/5 [==============================] - 6s 1s/step - loss: 851.1902 - r_square: -2.6707 - val_loss: 318.9502 - val_r_square: -0.6802 - lr: 1.0000e-04
Epoch 27/30
5/5 [==============================] - ETA: 0s - loss: 1001.7164 - r_square: -2.5450
Epoch 27: val_loss improved from 318.95020 to 265.62354, saving model to weights-27-265.62.hdf5
5/5 [==============================] - 6s 1s/step - loss: 1001.7164 - r_square: -2.5450 - val_loss: 265.6235 - val_r_square: -0.3993 - lr: 1.0000e-04
Epoch 28/30
5/5 [==============================] - ETA: 0s - loss: 429.7431 - r_square: -0.9815
Epoch 28: val_loss improved from 265.62354 to 232.07446, saving model to weights-28-232.07.hdf5
5/5 [==============================] - 6s 1s/step - loss: 429.7431 - r_square: -0.9815 - val_loss: 232.0745 - val_r_square: -0.2225 - lr: 1.0000e-04
Epoch 29/30
5/5 [==============================] - ETA: 0s - loss: 744.0077 - r_square: -1.6454
Epoch 29: val_loss improved from 232.07446 to 206.98068, saving model to weights-29-206.98.hdf5
5/5 [==============================] - 6s 1s/step - loss: 744.0077 - r_square: -1.6454 - val_loss: 206.9807 - val_r_square: -0.0903 - lr: 1.0000e-04
Epoch 30/30
5/5 [==============================] - ETA: 0s - loss: 721.0552 - r_square: -2.0117
Epoch 30: val_loss improved from 206.98068 to 193.21121, saving model to weights-30-193.21.hdf5
5/5 [==============================] - 6s 1s/step - loss: 721.0552 - r_square: -2.0117 - val_loss: 193.2112 - val_r_square: -0.0178 - lr: 1.0000e-04


# copy it there
## Copying the final model weights and the model arch JSON files
!cp weights-30-193.21.hdf5 /content/drive/MyDrive/CGM_Assignment
!cp resnet_50b.json /content/drive/MyDrive/CGM_Assignment


# Storing the TRAIN & VAL Losses
train_losses = history2.history['loss']
validation_losses = history2.history['val_loss']

# Generate a sequence of integers to represent the epoch numbers
epochs = range(1, 31)

with plt.style.context('seaborn-poster'):
    plt.figure(figsize=(18,8))
    # Plot and label the training and validation loss values
    plt.plot(epochs, train_losses, label='Training Loss', marker='o', markersize=10)
    plt.plot(epochs, validation_losses, label='Validation Loss', marker='o', markersize=10)

    # Add in a title, grid and axes labels
    plt.title('Training & Validation Loss', fontdict=title_font_dict)
    plt.xlabel('Epochs', fontdict=label_font_dict)
    plt.ylabel('Mean Squared Loss', fontdict=label_font_dict)
    plt.grid(which='major', linestyle=":", color='lightpink')
    plt.minorticks_on()

    # Set the tick locations
    plt.xticks(arange(0, 31, 1))

    # Display the plot
    plt.legend(loc='best')
    plt.show()


# Storing the TRAIN & VAL R-square values
train_r2 = history2.history['r_square']
val_r2 = history2.history['val_r_square']

# Generate a sequence of integers to represent the epoch numbers
epochs = range(1, 31)

with plt.style.context('seaborn-poster'):
    plt.figure(figsize=(18,8))
    # Plot and label the training and validation r-square values
    plt.plot(epochs, train_r2, label='Training R2', marker='^', markersize=10)
    plt.plot(epochs, val_r2, label='Validation R2', marker='^', markersize=10)
    plt.grid(which='major', linestyle=":", color='lightpink')
    plt.minorticks_on()

    # Add in a title, grid and axes labels
    plt.title('Training & Validation R-Square', fontdict=title_font_dict)
    plt.xlabel('Epochs', fontdict=label_font_dict)
    plt.ylabel('R-Square value', fontdict=label_font_dict)

    # Set the tick locations
    plt.xticks(arange(0, 31, 1))

    # Display the plot
    plt.legend(loc='best')
    plt.show()


# Generating the model predictions on the validation set
val_heights = resnet_50b.predict(X_val)
val_preds = []
for val in val_heights:
    val_preds.append((val[0]))

val_preds = np.array(val_preds)
val_preds

1/1 [==============================] - 2s 2s/step

array([74.21175 , 73.58505 , 73.8923  , 73.57336 , 73.87943 , 74.99591 ,
       73.747246, 73.07043 , 74.002716, 73.92571 ], dtype=float32)


# Actual VAL Tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# Validation R2 Score
r2_score(y_val, val_preds)

-0.07958032444093321


# Generating the model predictions on the TEST set
test_heights = resnet_50b.predict(X_test)
test_preds = []
for val in test_heights:
    test_preds.append((val[0]))

test_preds = np.array(test_preds)
test_preds

1/1 [==============================] - 1s 874ms/step

array([73.012566, 73.15798 ], dtype=float32)


# Actual TEST Tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# TEST R2 Score
r2_score(y_test, test_preds)

-9.485007287379972


# Fetching the directory name
curr_run_logdir2.split("/")[-1]

'run_2022_12_04-07_15_32'


# Any running instances
notebook.list()

Known TensorBoard instances:
  - port 6006: logdir logs (started 0:34:24 ago; pid 734)


# Running the TF Board
%tensorboard --logdir logs

Reusing TensorBoard on port 6006 (pid 734), started 0:34:27 ago. (Use '!kill 734' to kill it.)


# Generating the prediction on X_VAL
y_pred = resnet_50b.predict(X_val)

1/1 [==============================] - 0s 38ms/step


# Predictions on VAL set
val_preds = []
for val in y_pred:
    val_preds.append((val[0]))

val_preds = np.array(val_preds)
val_preds

array([74.21175 , 73.58505 , 73.8923  , 73.57336 , 73.87943 , 74.99591 ,
       73.747246, 73.07043 , 74.002716, 73.92571 ], dtype=float32)


# Actual VAL set tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# MSE on VAL Set
print("MSE on VAL set is --> {}".format(np.round(mean_squared_error(y_val, val_preds),3)))

MSE on VAL set is --> 189.26600646972656


# MAE on VAL Set
print("MAE on VAL set is --> {}".format(np.round(mean_absolute_error(y_val, val_preds),3)))

MAE on VAL set is --> 12.434000015258789


# MAPE on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(mean_absolute_percentage_error(y_val, val_preds),3)))

MAPE on VAL set is --> 0.1809999942779541


# R2 Score on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(r2_score(y_val, val_preds),3)))

MAPE on VAL set is --> -0.08


# Calculating the Error metrics
val_error =  np.abs(y_val - val_preds)
mean_error = np.mean(val_error)
min_error = np.min(val_error)
max_error = np.max(val_error)
std_error = np.std(val_error)


# Error metrics on VAL Set
print("####   Error Metrics on VAL Set   ####")
print("std_error: ",std_error)
print("mean_error: ",mean_error)
print("min_error: ",min_error)
print("max_error: ",max_error)

####   Error Metrics on VAL Set   ####
std_error:  5.887427
mean_error:  12.433993
min_error:  3.1266403
max_error:  25.897285


# Plotting the predictions
with plt.style.context('seaborn-poster'):
    # Visualising the Acutal and predicted Result
    plt.plot(y_val, color = 'lightgreen', label = 'Actual', marker='o')
    plt.plot(val_preds, color = 'coral', label = 'Predicted', marker='^')
    plt.grid(alpha = 0.3)
    plt.xlabel('Samples', fontdict=label_font_dict)
    plt.ylabel('Height in cm', fontdict=label_font_dict)
    plt.title('Actual vs Predicted', fontdict=title_font_dict)
    plt.legend()
    plt.show()


# Loading the model arch JSON file and model weights
!gdown 1YJWChf3MmiiKs86fpXww6_1jnoY68Jlb
!gdown 1sTuiWCHjKQJIjAHigapEFCZ1bKXqzJZw

Downloading...
From: https://drive.google.com/uc?id=1YJWChf3MmiiKs86fpXww6_1jnoY68Jlb
To: /content/resnet_50b.json
100% 86.6k/86.6k [00:00<00:00, 87.7MB/s]
Downloading...
From: https://drive.google.com/uc?id=1sTuiWCHjKQJIjAHigapEFCZ1bKXqzJZw
To: /content/weights-30-193.21.hdf5
100% 96.6M/96.6M [00:02<00:00, 41.4MB/s]


from tensorflow.keras.models import model_from_json


# loading Models' JSON and creating the Model
with open('resnet_50b.json', 'r') as model_json_file:
    loaded_model_json = model_json_file.read()
    
# Loading the model from json
loaded_model = model_from_json(loaded_model_json)

# load weights into new model
loaded_model.load_weights("weights-30-193.21.hdf5")


# Storing the DL Model with top layers excluded
new_model = tf.keras.models.Model(loaded_model.input, 
                                  loaded_model.get_layer('Flatten_for_hidden_layers').output)


# Generating the features for X_TRAIN
features_x_train = new_model.predict(X_train)

2/2 [==============================] - 12s 2s/step


# Shape of X_TRAIN features
features_x_train.shape

(53, 2048)


# Generating the X_VAL features
features_x_val = new_model.predict(X_val)

1/1 [==============================] - 1s 1s/step


# Shape of X_VAL features
features_x_val.shape

(10, 2048)


# Instantiating the Standard Scaler and training it
scaler = StandardScaler()
scaler.fit(features_x_train)

StandardScaler()


# Standardizing the X_TRAIN & X_VAL features
x_tr = scaler.transform(features_x_train)
x_vl = scaler.transform(features_x_val)


# Fitting the RF Regressor model with no HPTuning
clf = RandomForestRegressor(max_depth = 4, n_estimators = 10)
clf.fit(x_tr, y_train)

RandomForestRegressor(max_depth=4, n_estimators=10)


# Generating the prediction on X_VAL
y_pred = clf.predict(x_vl)


# Predictions on VAL set
y_pred

array([72.2599987 , 67.06999931, 64.46000023, 70.90999908, 64.05      ,
       97.6       , 70.58999901, 62.84000015, 99.90000153, 70.17999878])


# Actual VAL set tgt values
y_val

array([66.4, 59.1, 59.1, 76.7, 59.1, 86.4, 59.1, 66.4, 99.9, 63.2],
      dtype=float32)


# MSE on VAL Set
print("MSE on VAL set is --> {}".format(np.round(mean_squared_error(y_val, y_pred),3)))

MSE on VAL set is --> 50.347


# MAE on VAL Set
print("MAE on VAL set is --> {}".format(np.round(mean_absolute_error(y_val, y_pred),3)))

MAE on VAL set is --> 6.316


# MAPE on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(mean_absolute_percentage_error(y_val, y_pred),3)))

MAPE on VAL set is --> 0.096


# R2 Score on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(r2_score(y_val, y_pred),3)))

MAPE on VAL set is --> 0.713


# Calculating the Error metrics
val_error =  np.abs(y_val - y_pred)
mean_error = np.mean(val_error)
min_error = np.min(val_error)
max_error = np.max(val_error)
std_error = np.std(val_error)


# Error metrics on VAL Set
print("####   Error Metrics on VAL Set   ####")
print("std_error: ",std_error)
print("mean_error: ",mean_error)
print("min_error: ",min_error)
print("max_error: ",max_error)

####   Error Metrics on VAL Set   ####
std_error:  3.2334535792550008
mean_error:  6.315999755859375
min_error:  0.0
max_error:  11.490000534057614


# Plotting the predictions
with plt.style.context('seaborn-poster'):
    # Visualising the Acutal and predicted Result
    plt.plot(y_val, color = 'lightgreen', label = 'Actual', marker='o')
    plt.plot(y_pred, color = 'coral', label = 'Predicted', marker='^')
    plt.grid(alpha = 0.3)
    plt.xlabel('Samples', fontdict=label_font_dict)
    plt.ylabel('Height in cm', fontdict=label_font_dict)
    plt.title('Actual vs Predicted', fontdict=title_font_dict)
    plt.legend()
    plt.show()


# Save the model as a joblib
import joblib


# saving our model # model - model , filename-model_jlib
joblib.dump(clf , 'rf_model_jlib')

['rf_model_jlib']


# Copying the model into the Drive
!cp rf_model_jlib /content/drive/MyDrive/CGM_Assignment


# opening the file- model_jlib
loaded_rf = joblib.load('rf_model_jlib')


from sklearn.svm import SVR


# Seggregating the Features and TGT column
hp_dist_df = hp_df.drop(['Img_name'], axis=1)
X = hp_dist_df.drop(['Height'],axis=1)
y = hp_dist_df['Height']


hp_dist_df


# Train and Validation Split
x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=2022)


x_train.shape, x_test.shape, y_train.shape, y_test.shape

((7, 34), (2, 34), (7,), (2,))


# Instantiating the Standard Scaler and training it
scaler = StandardScaler()
scaler.fit(x_train)

StandardScaler()


# Standardizing the X_TRAIN & X_VAL features
x_tr = scaler.transform(x_train)
x_vl = scaler.transform(x_test)


# Instantiating the SVR model
svm_reg = SVR()

# Training the model
svm_reg_model = svm_reg.fit(x_tr, y_train)


# Generating the prediction on X_TRAIN
y_pred_tr = svm_reg_model.predict(x_tr)


# Predictions on TRAIN set
y_pred_tr

array([75.36824013, 75.17395236, 77.72584253, 77.86053893, 77.39585779,
       75.12385601, 76.7       ])


# Actual TRAIN set tgt values
y_train = y_train.values
y_train

array([ 59.1,  63.2,  86.4,  99.9, 106.1,  66.4,  76.7])


# MSE on TRAIN Set
print("MSE on TRAIN set is --> {}".format(np.round(mean_squared_error(y_train, y_pred_tr),3)))

MSE on TRAIN set is --> 267.006


# MAE on TRAIN Set
print("MAE on TRAIN set is --> {}".format(np.round(mean_absolute_error(y_train, y_pred_tr),3)))

MAE on TRAIN set is --> 13.769


# MAPE on TRAIN Set
print("MAPE on VAL set is --> {}".format(np.round(mean_absolute_percentage_error(y_train, y_pred_tr),3)))

MAPE on VAL set is --> 0.17


# R2 Score on TRAIN Set
print("R2 Score on TRAIN set is --> {}".format(np.round(r2_score(y_train, y_pred_tr),3)))

R2 Score on TRAIN set is --> 0.08


# Calculating the Error metrics
train_error =  np.abs(y_train - y_pred_tr)
mean_error = np.mean(train_error)
min_error = np.min(train_error)
max_error = np.max(train_error)
std_error = np.std(train_error)


# Error metrics on TRAIN Set
print("######   Error Metrics on TRAIN Set   ######\n")
print("-- std_error: {}\n".format(np.round(std_error,3)))
print("-- mean_error: {}\n".format(np.round(mean_error,3)))
print("-- min_error: {}\n".format(np.round(min_error,3)))
print("-- max_error: {}\n".format(np.round(max_error,3)))

######   Error Metrics on TRAIN Set   ######

-- std_error: 8.799

-- mean_error: 13.769

-- min_error: 0.0

-- max_error: 28.704


# Plotting the predictions
with plt.style.context('seaborn-poster'):
    # Visualising the Acutal and predicted Result
    plt.plot(y_train, color = 'lightgreen', label = 'Actual', marker='o')
    plt.plot(y_pred_tr, color = 'coral', label = 'Predicted', marker='^')
    plt.grid(alpha = 0.3)
    plt.xlabel('Samples', fontdict=label_font_dict)
    plt.ylabel('Height in cm', fontdict=label_font_dict)
    plt.title('Actual vs Predicted', fontdict=title_font_dict)
    plt.legend()
    plt.show()


# Generating the prediction on X_VAL
y_pred = svm_reg_model.predict(x_vl)


# Predictions on VAL set
y_pred

array([77.10322727, 76.65127704])


# Actual VAL set tgt values
y_test = y_test.values
y_test

array([100.6,  87.1])


# MSE on VAL Set
print("MSE on VAL set is --> {}".format(np.round(mean_squared_error(y_test, y_pred),3)))

MSE on VAL set is --> 330.637


# MAE on VAL Set
print("MAE on VAL set is --> {}".format(np.round(mean_absolute_error(y_test, y_pred),3)))

MAE on VAL set is --> 16.973


# MAPE on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(mean_absolute_percentage_error(y_test, y_pred),3)))

MAPE on VAL set is --> 0.177


# R2 Score on VAL Set
print("MAPE on VAL set is --> {}".format(np.round(r2_score(y_test, y_pred),3)))

MAPE on VAL set is --> -6.257


# Calculating the Error metrics
val_error =  np.abs(y_test - y_pred)
mean_error = np.mean(val_error)
min_error = np.min(val_error)
max_error = np.max(val_error)
std_error = np.std(val_error)


# Error metrics on TRAIN Set
print("######   Error Metrics on TRAIN Set   ######\n")
print("-- std_error: {}\n".format(np.round(std_error,3)))
print("-- mean_error: {}\n".format(np.round(mean_error,3)))
print("-- min_error: {}\n".format(np.round(min_error,3)))
print("-- max_error: {}\n".format(np.round(max_error,3)))

######   Error Metrics on TRAIN Set   ######

-- std_error: 6.524

-- mean_error: 16.973

-- min_error: 10.449

-- max_error: 23.497


# Plotting the predictions
with plt.style.context('seaborn-poster'):
    # Visualising the Acutal and predicted Result
    plt.plot(y_test, color = 'lightgreen', label = 'Actual', marker='o')
    plt.plot(y_pred, color = 'coral', label = 'Predicted', marker='^')
    plt.grid(alpha = 0.3)
    plt.xlabel('Samples', fontdict=label_font_dict)
    plt.ylabel('Height in cm', fontdict=label_font_dict)
    plt.title('Actual vs Predicted', fontdict=title_font_dict)
    # Set the tick locations
    plt.xticks(arange(0, 2, 1))
    plt.legend()
    plt.show()

	Img_name	nose_top_dist	nose_bottom_dist	left_eye_top_dist	left_eye_bottom_dist	right_eye_top_dist	right_eye_bottom_dist	left_ear_top_dist	left_ear_bottom_dist	right_ear_top_dist	right_ear_bottom_dist	left_shoulder_top_dist	left_shoulder_bottom_dist	right_shoulder_top_dist	right_shoulder_bottom_dist	left_elbow_top_dist	left_elbow_bottom_dist	right_elbow_top_dist	right_elbow_bottom_dist	left_wrist_top_dist	left_wrist_bottom_dist	right_wrist_top_dist	right_wrist_bottom_dist	left_hip_top_dist	left_hip_bottom_dist	right_hip_top_dist	right_hip_bottom_dist	left_knee_top_dist	left_knee_bottom_dist	right_knee_top_dist	right_knee_bottom_dist	left_ankle_top_dist	left_ankle_bottom_dist	right_ankle_top_dist	right_ankle_bottom_dist	Height
0	/content/Dataset/1d5fee00-ada4-11eb-a80b-3f50a...	139.691	453.810	151.440	460.771	113.032	476.286	180.259	445.126	101.246	478.053	224.946	388.402	163.677	412.205	291.861	332.618	218.935	368.847	319.802	276.817	235.590	342.005	352.147	229.775	330.627	247.948	457.988	118.041	400.249	210.554	426.089	152.056	458.407	194.857	76.7
1	/content/Dataset/1eddf4c0-ad9f-11eb-a80b-ef4ad...	219.240	376.907	107.062	464.210	208.241	395.238	139.040	432.574	202.198	386.158	170.682	399.706	237.697	349.583	130.723	440.394	287.228	340.067	95.316	481.054	291.823	375.793	306.436	272.108	332.235	238.045	374.111	226.972	393.302	177.855	447.702	165.827	464.532	105.850	66.4
2	/content/Dataset/38e29bc0-8fc0-11eb-b5f2-0742d...	114.684	464.978	121.393	460.750	209.028	415.961	139.339	438.516	210.612	412.764	180.103	396.358	244.709	353.241	117.177	462.027	254.459	392.036	107.047	483.412	254.249	442.271	334.531	250.052	365.415	211.029	409.651	202.716	429.890	146.674	489.473	148.322	489.456	103.502	63.2
3	/content/Dataset/6cd5d480-8b8b-11eb-b152-f7b11...	150.096	505.597	155.437	516.805	127.744	524.638	185.402	504.764	125.420	520.907	242.182	436.003	178.191	464.880	308.353	359.789	245.235	406.229	362.151	296.329	311.701	352.524	369.434	274.015	348.848	301.257	464.828	182.184	445.466	225.565	543.030	128.115	531.762	177.651	87.1
4	/content/Dataset/76ed6fc0-8e1d-11eb-ad70-7be1a...	91.655	500.699	88.932	502.501	89.834	501.079	122.090	470.116	167.362	470.016	160.697	430.627	194.623	439.693	129.537	468.497	212.710	471.281	84.680	507.193	183.968	515.070	308.260	285.770	327.234	268.397	381.836	239.404	422.678	168.536	479.981	165.535	507.097	90.993	59.1
5	/content/Dataset/8d23fa60-9201-11eb-b3e6-cbfb4...	190.423	472.569	179.159	487.789	172.724	488.188	131.322	488.105	136.639	483.716	169.193	435.949	209.694	408.771	239.258	375.096	310.914	304.460	271.681	333.434	289.150	337.040	329.403	282.248	353.680	251.499	447.742	172.448	467.127	140.664	547.518	117.861	567.176	64.311	86.4
6	/content/Dataset/9dc3eba0-8d8d-11eb-8d3f-4776b...	142.212	458.133	136.621	473.452	116.796	480.265	155.741	472.965	103.867	489.563	221.702	394.308	180.226	415.960	300.493	308.431	263.732	342.875	364.402	238.180	337.507	279.494	329.515	265.448	312.568	285.209	422.124	171.750	414.752	195.807	488.571	113.838	492.155	138.798	100.6
7	/content/Dataset/9dd020a0-8d8d-11eb-8d3f-bfd9f...	118.411	437.884	110.964	454.636	96.762	457.973	135.269	443.555	101.644	451.652	199.977	375.700	158.455	396.762	280.915	285.278	247.099	316.309	344.566	216.070	329.922	244.375	325.652	228.244	312.485	246.740	409.378	146.260	405.849	163.278	479.062	91.341	475.298	110.771	106.1
8	/content/Dataset/d3757800-8dff-11eb-a328-472d4...	179.026	501.837	174.395	509.670	176.691	516.647	111.132	530.799	170.003	517.505	173.186	468.816	238.792	433.049	277.820	377.785	333.128	325.585	348.933	315.665	400.402	253.969	354.905	290.880	378.720	261.624	464.386	197.124	471.840	167.764	571.270	159.135	583.359	62.913	99.9

	nose_top_dist	nose_bottom_dist	left_eye_top_dist	left_eye_bottom_dist	right_eye_top_dist	right_eye_bottom_dist	left_ear_top_dist	left_ear_bottom_dist	right_ear_top_dist	right_ear_bottom_dist	left_shoulder_top_dist	left_shoulder_bottom_dist	right_shoulder_top_dist	right_shoulder_bottom_dist	left_elbow_top_dist	left_elbow_bottom_dist	right_elbow_top_dist	right_elbow_bottom_dist	left_wrist_top_dist	left_wrist_bottom_dist	right_wrist_top_dist	right_wrist_bottom_dist	left_hip_top_dist	left_hip_bottom_dist	right_hip_top_dist	right_hip_bottom_dist	left_knee_top_dist	left_knee_bottom_dist	right_knee_top_dist	right_knee_bottom_dist	left_ankle_top_dist	left_ankle_bottom_dist	right_ankle_top_dist	right_ankle_bottom_dist	Height
0	139.691	453.810	151.440	460.771	113.032	476.286	180.259	445.126	101.246	478.053	224.946	388.402	163.677	412.205	291.861	332.618	218.935	368.847	319.802	276.817	235.590	342.005	352.147	229.775	330.627	247.948	457.988	118.041	400.249	210.554	426.089	152.056	458.407	194.857	76.7
1	219.240	376.907	107.062	464.210	208.241	395.238	139.040	432.574	202.198	386.158	170.682	399.706	237.697	349.583	130.723	440.394	287.228	340.067	95.316	481.054	291.823	375.793	306.436	272.108	332.235	238.045	374.111	226.972	393.302	177.855	447.702	165.827	464.532	105.850	66.4
2	114.684	464.978	121.393	460.750	209.028	415.961	139.339	438.516	210.612	412.764	180.103	396.358	244.709	353.241	117.177	462.027	254.459	392.036	107.047	483.412	254.249	442.271	334.531	250.052	365.415	211.029	409.651	202.716	429.890	146.674	489.473	148.322	489.456	103.502	63.2
3	150.096	505.597	155.437	516.805	127.744	524.638	185.402	504.764	125.420	520.907	242.182	436.003	178.191	464.880	308.353	359.789	245.235	406.229	362.151	296.329	311.701	352.524	369.434	274.015	348.848	301.257	464.828	182.184	445.466	225.565	543.030	128.115	531.762	177.651	87.1
4	91.655	500.699	88.932	502.501	89.834	501.079	122.090	470.116	167.362	470.016	160.697	430.627	194.623	439.693	129.537	468.497	212.710	471.281	84.680	507.193	183.968	515.070	308.260	285.770	327.234	268.397	381.836	239.404	422.678	168.536	479.981	165.535	507.097	90.993	59.1
5	190.423	472.569	179.159	487.789	172.724	488.188	131.322	488.105	136.639	483.716	169.193	435.949	209.694	408.771	239.258	375.096	310.914	304.460	271.681	333.434	289.150	337.040	329.403	282.248	353.680	251.499	447.742	172.448	467.127	140.664	547.518	117.861	567.176	64.311	86.4
6	142.212	458.133	136.621	473.452	116.796	480.265	155.741	472.965	103.867	489.563	221.702	394.308	180.226	415.960	300.493	308.431	263.732	342.875	364.402	238.180	337.507	279.494	329.515	265.448	312.568	285.209	422.124	171.750	414.752	195.807	488.571	113.838	492.155	138.798	100.6
7	118.411	437.884	110.964	454.636	96.762	457.973	135.269	443.555	101.644	451.652	199.977	375.700	158.455	396.762	280.915	285.278	247.099	316.309	344.566	216.070	329.922	244.375	325.652	228.244	312.485	246.740	409.378	146.260	405.849	163.278	479.062	91.341	475.298	110.771	106.1
8	179.026	501.837	174.395	509.670	176.691	516.647	111.132	530.799	170.003	517.505	173.186	468.816	238.792	433.049	277.820	377.785	333.128	325.585	348.933	315.665	400.402	253.969	354.905	290.880	378.720	261.624	464.386	197.124	471.840	167.764	571.270	159.135	583.359	62.913	99.9

`CGM - Assignment` : Predicting Height of the Children 👧👦📏¶

`Examples of various measurements` are shown below:¶

`NOTEBOOK Description` 📗¶

`NOTEBOOK Contents` ✍️¶

⭐ `Image Augmentations`¶

⭐ `Performance Metrics`¶

⭐ `Models Strategies`¶

Import_Libraries¶

Dataset_Import¶

`Data_Preparation`¶

Reading the images¶

Image Data Generator¶

`Image_Augmentation`¶

`Defining_Performance_Metrics`¶

`Models_Training_Configuration`¶

`Models`¶

`A.Fine_Tuning---Last_3_Layers+Custom_TopLayers---ImageNet_Weights`¶

`A1.ResNet---50`¶

`Plot the Loss and R2 Graphs`¶

`LOSS_Curves-1`¶

`R_Square_Curves-1`¶

Evaluating the Model¶

`VAL_Evaluation-1`¶

`TEST_Evaluation-1`¶

Running TensorBoard - A1¶

`B.Fine_Tuning---Custom_TopLayers---ImageNet_Weights`¶

`B1.ResNet---50`¶

`LOSS_Curves-2`¶

`R_Square_Curves-2`¶

Evaluating the Model¶

`VAL_Evaluation-2`¶

`TEST_Evaluation-2`¶

Running TensorBoard - B1¶

`Predictions_and_Scores_on_VAL_Set`¶

`C.DL_FEATURES---ML_Regressor_Model`¶

`Predictions_and_Scores_on_VAL_Set`¶

`D.TOP_and_BOTTOM_DistanceFeatures---ML_Regressor_Model`¶

`Predictions_and_Scores_on_TRAIN_Set`¶

`Predictions_and_Scores_on_VAL_Set`¶

SUMMARY¶

CGM - Assignment : Predicting Height of the Children 👧👦📏¶

Examples of various measurements are shown below:¶

NOTEBOOK Description 📗¶

NOTEBOOK Contents ✍️¶

⭐ Image Augmentations¶

⭐ Performance Metrics¶

⭐ Models Strategies¶

Import_Libraries¶

Dataset_Import¶

Data_Preparation¶

Reading the images¶

Image Data Generator¶

Image_Augmentation¶

Defining_Performance_Metrics¶

Models_Training_Configuration¶

Models¶

A.Fine_Tuning---Last_3_Layers__+__Custom_TopLayers---ImageNet_Weights¶

A1.ResNet---50¶

Plot the Loss and R2 Graphs¶

LOSS_Curves-1¶

R_Square_Curves-1¶

Evaluating the Model¶

VAL_Evaluation-1¶

TEST_Evaluation-1¶

Running TensorBoard - A1¶

B.Fine_Tuning---Custom_TopLayers---ImageNet_Weights¶

B1.ResNet---50¶

LOSS_Curves-2¶

R_Square_Curves-2¶

Evaluating the Model¶

VAL_Evaluation-2¶

TEST_Evaluation-2¶

Running TensorBoard - B1¶

Predictions_and_Scores_on_VAL_Set¶

C.DL_FEATURES---ML_Regressor_Model¶

Predictions_and_Scores_on_VAL_Set¶

D.TOP_and_BOTTOM_DistanceFeatures---ML_Regressor_Model¶

Predictions_and_Scores_on_TRAIN_Set¶

Predictions_and_Scores_on_VAL_Set¶

SUMMARY¶

`CGM - Assignment` : Predicting Height of the Children 👧👦📏¶

`Examples of various measurements` are shown below:¶

`NOTEBOOK Description` 📗¶

`NOTEBOOK Contents` ✍️¶

⭐ `Image Augmentations`¶

⭐ `Performance Metrics`¶

⭐ `Models Strategies`¶

`Data_Preparation`¶

`Image_Augmentation`¶

`Defining_Performance_Metrics`¶

`Models_Training_Configuration`¶

`Models`¶

`A.Fine_Tuning---Last_3_Layers+Custom_TopLayers---ImageNet_Weights`¶

`A1.ResNet---50`¶

`Plot the Loss and R2 Graphs`¶

`LOSS_Curves-1`¶

`R_Square_Curves-1`¶

`VAL_Evaluation-1`¶

`TEST_Evaluation-1`¶

`B.Fine_Tuning---Custom_TopLayers---ImageNet_Weights`¶

`B1.ResNet---50`¶

`LOSS_Curves-2`¶

`R_Square_Curves-2`¶

`VAL_Evaluation-2`¶

`TEST_Evaluation-2`¶

`Predictions_and_Scores_on_VAL_Set`¶

`C.DL_FEATURES---ML_Regressor_Model`¶

`Predictions_and_Scores_on_VAL_Set`¶

`D.TOP_and_BOTTOM_DistanceFeatures---ML_Regressor_Model`¶

`Predictions_and_Scores_on_TRAIN_Set`¶

`Predictions_and_Scores_on_VAL_Set`¶